From bb7e8d682ed8bbc8f11fd5251b5d334113676a0a Mon Sep 17 00:00:00 2001 From: jackwener Date: Thu, 20 Aug 2026 15:56:06 +0800 Subject: [PATCH] fix(models): resolve reasoning capabilities from provider facts --- .github/workflows/model-metadata-sync.yml | 72 + package.json | 2 +- .../core/src/__tests__/model-metadata.test.ts | 4 +- .../core/src/__tests__/model-thinking.test.ts | 29 + .../__tests__/runtime-policy-codec.test.ts | 16 +- packages/core/src/llm-connections.ts | 5 + packages/core/src/model-metadata.generated.ts | 2187 ++++++++++++++--- packages/core/src/model-metadata.ts | 15 +- packages/core/src/model-thinking.ts | 27 +- packages/core/src/provider-registry.ts | 35 +- .../connection-catalog-codec.ts | 36 +- .../__tests__/model-factory-thinking.test.ts | 22 +- .../__tests__/provider-contract-overrides.ts | 3 + packages/runtime/src/model-factory.ts | 26 +- packages/runtime/src/model-fetcher.ts | 3 + .../src/telemetry/model-pricing.generated.ts | 630 ++++- scripts/sync-model-metadata.mjs | 1 + 17 files changed, 2607 insertions(+), 506 deletions(-) create mode 100644 .github/workflows/model-metadata-sync.yml diff --git a/.github/workflows/model-metadata-sync.yml b/.github/workflows/model-metadata-sync.yml new file mode 100644 index 0000000000..bbda349464 --- /dev/null +++ b/.github/workflows/model-metadata-sync.yml @@ -0,0 +1,72 @@ +name: Model metadata sync + +on: + schedule: + # Keep newly released models from waiting for an unrelated code change. + - cron: '43 4 * * *' + workflow_dispatch: + +permissions: + contents: write + pull-requests: write + +concurrency: + group: model-metadata-sync + cancel-in-progress: true + +jobs: + sync: + runs-on: ubuntu-latest + timeout-minutes: 10 + steps: + - name: Check out the repository + uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + with: + fetch-depth: 0 + + - name: Set up Node.js + uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020 # v7.0.0 + with: + node-version: '24' + + - name: Install validation dependencies + run: npm ci --ignore-scripts + + - name: Refresh models.dev facts + run: npm run sync:model-metadata + + - id: changes + name: Detect catalog changes + run: | + if git diff --quiet; then + echo "changed=false" >> "$GITHUB_OUTPUT" + else + echo "changed=true" >> "$GITHUB_OUTPUT" + fi + + - name: Validate generated catalog contracts + if: steps.changes.outputs.changed == 'true' + run: | + npm --workspace @maka/core run build + node -e "import('./packages/core/dist/provider-registry.js')" + + - name: Open or update the sync pull request + if: steps.changes.outputs.changed == 'true' + env: + GH_TOKEN: ${{ github.token }} + BRANCH: automation/models-dev-sync + run: | + git config user.name github-actions[bot] + git config user.email 41898282+github-actions[bot]@users.noreply.github.com + git switch -C "$BRANCH" + git add packages/core/src/model-metadata.generated.ts packages/runtime/src/telemetry/model-pricing.generated.ts + git commit -m "chore(models): refresh models.dev metadata" + git push --force origin "HEAD:$BRANCH" + + if [[ "$(gh pr list --head "$BRANCH" --state open --json number --jq length)" == "0" ]]; then + gh pr create \ + --base main \ + --head "$BRANCH" \ + --title "chore(models): refresh models.dev metadata" \ + --body "Automated refresh of bundled model capabilities and pricing from models.dev." + fi diff --git a/package.json b/package.json index d34dd504e1..8f42a994a6 100644 --- a/package.json +++ b/package.json @@ -58,7 +58,7 @@ "astryx:theme": "node scripts/build-astryx-theme.mjs", "astryx:surface-inventory": "node scripts/check-astryx-surface-inventory.mjs", "astryx:surface-inventory:write": "node scripts/generate-astryx-surface-inventory.mjs", - "sync:model-metadata": "node scripts/sync-model-metadata.mjs", + "sync:model-metadata": "node scripts/sync-model-metadata.mjs && biome format --write --files-max-size=5000000 packages/core/src/model-metadata.generated.ts packages/runtime/src/telemetry/model-pricing.generated.ts", "generate:bundled-skills": "node scripts/gen-bundled-skill-catalog.mjs", "cost:deepseek-baseline": "node scripts/deepseek-live-cost-baseline.mjs", "computer-use": "node scripts/computer-use.mjs", diff --git a/packages/core/src/__tests__/model-metadata.test.ts b/packages/core/src/__tests__/model-metadata.test.ts index 9bf9ae6322..416a1df1da 100644 --- a/packages/core/src/__tests__/model-metadata.test.ts +++ b/packages/core/src/__tests__/model-metadata.test.ts @@ -64,9 +64,11 @@ describe('openAiAdapterApiProtocol', () => { assert.equal(openAiAdapterApiProtocol('gpt-4o'), 'openai-chat'); }); - it('routes only xAI Grok 4.5 through Responses', () => { + it('routes xAI Responses-generation Grok models by family contract', () => { assert.equal(openAiAdapterApiProtocol('grok-4.5', 'xai'), 'openai-responses'); assert.equal(openAiAdapterApiProtocol('grok-4.5', 'xai-oauth'), 'openai-responses'); + assert.equal(openAiAdapterApiProtocol('grok-4.6', 'xai'), 'openai-responses'); + assert.equal(openAiAdapterApiProtocol('grok-5', 'xai'), 'openai-responses'); assert.equal(openAiAdapterApiProtocol('grok-4.3', 'xai'), 'openai-chat'); assert.equal(openAiAdapterApiProtocol('grok-4.5', 'openai'), 'openai-chat'); }); diff --git a/packages/core/src/__tests__/model-thinking.test.ts b/packages/core/src/__tests__/model-thinking.test.ts index a136a28808..656e51b4ec 100644 --- a/packages/core/src/__tests__/model-thinking.test.ts +++ b/packages/core/src/__tests__/model-thinking.test.ts @@ -7,10 +7,39 @@ import { resolveThinkingLevel, deriveThinkingChoices, thinkingOptionsForModel, + thinkingOptionsForConnection, thinkingVariantsForConnection, thinkingVariantsForModel, } from '../model-thinking.js'; +test('provider inventory reasoning controls outrank the bundled snapshot', () => { + const connection = { + providerType: 'github-copilot', + models: [ + { + id: 'grok-4.6', + capabilities: { reasoning: true }, + thinkingOptions: { efforts: ['low', 'xhigh'] }, + }, + ], + } as const; + assert.deepEqual(thinkingOptionsForConnection(connection, 'grok-4.6'), { + efforts: ['low', 'xhigh'], + }); + assert.deepEqual(thinkingVariantsForConnection(connection, 'grok-4.6'), ['low', 'xhigh']); + + assert.equal( + thinkingOptionsForConnection( + { + providerType: 'xai', + models: [{ id: 'grok-4.5', capabilities: { reasoning: false } }], + }, + 'grok-4.5', + ), + undefined, + ); +}); + test('declarable relay levels are every intensity tier but off', () => { // `off` is a disable-wire encoding (reasoning_effort 'none'), not an // intensity tier — a hybrid UI/data contract keeps it out of declarations. diff --git a/packages/core/src/__tests__/runtime-policy-codec.test.ts b/packages/core/src/__tests__/runtime-policy-codec.test.ts index e39d69eb08..2732698405 100644 --- a/packages/core/src/__tests__/runtime-policy-codec.test.ts +++ b/packages/core/src/__tests__/runtime-policy-codec.test.ts @@ -383,12 +383,24 @@ test('relay model profiles round-trip canonical entries and drafts, strictly', ( test('normalizes exact bounded model discovery results', () => { assert.deepEqual( normalizeConnectionModelDiscoveryResult({ - models: [{ id: 'gpt-5', capabilities: { chat: true } }], + models: [ + { + id: 'gpt-5', + capabilities: { chat: true }, + thinkingOptions: { efforts: ['low', 'high'] }, + }, + ], source: 'fetched', fetchedAt: 42, }), { - models: [{ id: 'gpt-5', capabilities: { chat: true } }], + models: [ + { + id: 'gpt-5', + capabilities: { chat: true }, + thinkingOptions: { efforts: ['low', 'high'] }, + }, + ], source: 'fetched', fetchedAt: 42, }, diff --git a/packages/core/src/llm-connections.ts b/packages/core/src/llm-connections.ts index 5f0cf88707..ca3728b12b 100644 --- a/packages/core/src/llm-connections.ts +++ b/packages/core/src/llm-connections.ts @@ -81,6 +81,11 @@ export interface ModelInfo { /** Provider-hosted live web search, using this exact model and connection. */ webSearch?: boolean; }; + /** Exact reasoning controls advertised by this provider's model inventory. */ + thinkingOptions?: { + efforts?: string[]; + toggle?: boolean; + }; /** Multimodal input/output support from provider catalog metadata. */ modalities?: { input: Array<'text' | 'image' | 'audio' | 'pdf'>; diff --git a/packages/core/src/model-metadata.generated.ts b/packages/core/src/model-metadata.generated.ts index cc8472d647..42218a83b3 100644 --- a/packages/core/src/model-metadata.generated.ts +++ b/packages/core/src/model-metadata.generated.ts @@ -234,6 +234,35 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, }, alibaba: { + 'deepseek-v4-flash-0731': { + displayName: 'DeepSeek V4 Flash 0731', + description: + 'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding', + lifecycle: 'active', + docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models', + contextWindow: 1000000, + maxOutputTokens: 384000, + knowledgeCutoff: '2025-05', + structuredOutput: true, + lastUpdated: '2026-07-31', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'glm-5.2': { + displayName: 'GLM-5.2', + description: + 'Open flagship GLM for long-horizon coding agents and million-token context work', + lifecycle: 'active', + docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models', + contextWindow: 1000000, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-06-13', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'xhigh', 'max'] }, + modalities: { input: ['text'], output: ['text'] }, + }, 'qvq-max': { displayName: 'QVQ Max', description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', @@ -893,6 +922,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 202752, maxOutputTokens: 16384, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-12-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -905,6 +935,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan', contextWindow: 202752, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-02-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -918,6 +949,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2026-01-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -930,6 +962,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan', contextWindow: 196608, maxOutputTokens: 24576, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -942,6 +975,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2026-02-03', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -954,6 +988,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-07-23', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -966,6 +1001,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-01-23', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -978,6 +1014,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-02-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1004,6 +1041,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1030,6 +1068,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 64000, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-06-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1045,6 +1084,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 202752, maxOutputTokens: 16384, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-12-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1057,6 +1097,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan', contextWindow: 202752, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-02-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1070,6 +1111,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2026-01-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1083,6 +1125,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 196608, inputLimit: 196601, maxOutputTokens: 24576, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -1095,6 +1138,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2026-02-03', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -1107,6 +1151,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-07-23', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -1119,6 +1164,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-01-23', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -1131,6 +1177,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-02-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1157,6 +1204,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1183,6 +1231,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 64000, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-06-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1199,6 +1248,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 65536, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2025-12-05', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1213,9 +1263,10 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-24', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { toggle: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'deepseek-v4-flash-0731': { @@ -1228,6 +1279,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-07-31', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'], toggle: true }, @@ -1242,9 +1294,25 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-24', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { toggle: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'deepseek-v4-pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', + contextWindow: 1000000, + maxOutputTokens: 384000, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'glm-5': { @@ -1256,6 +1324,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 202752, maxOutputTokens: 16384, structuredOutput: true, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1270,6 +1339,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 202752, maxOutputTokens: 128000, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-07', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1284,6 +1354,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'] }, @@ -1296,6 +1367,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 0, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-07-17', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['image', 'text'], output: [] }, @@ -1307,6 +1379,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 0, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-07-17', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['image', 'text'], output: [] }, @@ -1318,6 +1391,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 0, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-07-17', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: [] }, @@ -1332,6 +1406,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 98304, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-01', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1346,6 +1421,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-21', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1361,6 +1437,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-12', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1374,6 +1451,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 196608, inputLimit: 196601, maxOutputTokens: 32768, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -1385,6 +1463,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-03-03', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1396,6 +1475,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-03-03', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1408,6 +1488,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1422,6 +1503,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 65536, knowledgeCutoff: '2025-04', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1436,6 +1518,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-05-21', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1450,6 +1533,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 65536, knowledgeCutoff: '2025-04', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1464,6 +1548,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-08-03', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true }, @@ -1478,6 +1563,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-07-19', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] }, @@ -1490,6 +1576,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-05-29', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1501,6 +1588,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-05-29', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1516,6 +1604,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 65536, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2025-12-05', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1530,6 +1619,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-24', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'], toggle: true }, @@ -1545,6 +1635,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-07-31', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'], toggle: true }, @@ -1559,11 +1650,27 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-24', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek-v4-pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', + contextWindow: 1000000, + maxOutputTokens: 384000, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'glm-5': { displayName: 'GLM-5', description: @@ -1573,6 +1680,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 202752, maxOutputTokens: 16384, structuredOutput: true, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1587,6 +1695,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 202752, maxOutputTokens: 128000, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-07', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1601,6 +1710,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'] }, @@ -1613,6 +1723,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 0, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-07-17', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['image', 'text'], output: [] }, @@ -1624,6 +1735,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 0, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-07-17', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['image', 'text'], output: [] }, @@ -1635,6 +1747,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 0, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-07-17', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: [] }, @@ -1649,6 +1762,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 98304, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-01', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1663,6 +1777,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-21', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1678,6 +1793,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-12', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1691,6 +1807,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 196608, inputLimit: 196601, maxOutputTokens: 32768, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -1702,6 +1819,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-03-03', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1713,6 +1831,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-03-03', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1725,6 +1844,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1739,6 +1859,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 65536, knowledgeCutoff: '2025-04', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1753,6 +1874,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-05-21', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1767,6 +1889,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 65536, knowledgeCutoff: '2025-04', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -1781,6 +1904,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-08-03', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true }, @@ -1795,6 +1919,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-07-19', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] }, @@ -1807,6 +1932,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-05-29', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1818,6 +1944,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview', contextWindow: 8192, maxOutputTokens: 0, + isFree: true, lastUpdated: '2026-05-29', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -1850,19 +1977,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text'], output: ['text'] }, }, - 'zai-glm-4.7': { - displayName: 'Z.AI GLM-4.7', - description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering', - lifecycle: 'beta', - docsUrl: 'https://inference-docs.cerebras.ai/models/overview', - contextWindow: 131072, - maxOutputTokens: 40960, - structuredOutput: true, - lastUpdated: '2026-06-10', - capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['none'] }, - modalities: { input: ['text'], output: ['text'] }, - }, }, cohere: { 'c4ai-aya-expanse-32b': { @@ -2035,6 +2149,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 64000, knowledgeCutoff: '2025-09-23', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-09', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['none', 'high'] }, @@ -2068,6 +2183,35 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: true, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, }, + '@cf/deepseek-ai/deepseek-v4-flash-0731': { + displayName: 'DeepSeek V4 Flash 0731', + description: + 'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding', + lifecycle: 'active', + docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', + contextWindow: 1310720, + maxOutputTokens: 1048576, + knowledgeCutoff: '2025-05', + structuredOutput: true, + lastUpdated: '2026-07-31', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + '@cf/deepseek-ai/deepseek-v4-pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', + contextWindow: 1048576, + maxOutputTokens: 1048576, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, '@cf/google/gemma-4-26b-a4b-it': { displayName: 'Gemma 4 26B A4B IT', description: 'Open Gemma instruction model for efficient chat and self-hosted deployments', @@ -2100,8 +2244,9 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', contextWindow: 32000, maxOutputTokens: 32000, + knowledgeCutoff: '2023-12', structuredOutput: false, - lastUpdated: '2024-07-25', + lastUpdated: '2024-07-23', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, }, @@ -2112,6 +2257,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', contextWindow: 128000, maxOutputTokens: 128000, + knowledgeCutoff: '2023-12', structuredOutput: false, lastUpdated: '2024-09-25', capabilities: { vision: true, reasoning: false, functionCalling: false }, @@ -2177,8 +2323,9 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', contextWindow: 131072, maxOutputTokens: 131072, + knowledgeCutoff: '2023-12', structuredOutput: false, - lastUpdated: '2025-01-22', + lastUpdated: '2024-07-23', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, }, @@ -2271,7 +2418,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 32768, maxOutputTokens: 32768, structuredOutput: false, - lastUpdated: '2025-02-27', + lastUpdated: '2024-11-12', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, }, @@ -2283,10 +2430,23 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 32768, maxOutputTokens: 32768, structuredOutput: false, - lastUpdated: '2025-04-30', + lastUpdated: '2025-04-28', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + '@cf/qwen/qwen3.8-27b': { + displayName: 'Qwen3.8 27B', + description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', + contextWindow: 262144, + maxOutputTokens: 262144, + structuredOutput: true, + lastUpdated: '2026-08-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, '@cf/qwen/qwq-32b': { displayName: 'Qwq 32B', description: 'Qwen reasoning model for deliberate problem solving, math, and coding', @@ -2294,6 +2454,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', contextWindow: 24000, maxOutputTokens: 24000, + knowledgeCutoff: '2024-04', structuredOutput: false, lastUpdated: '2025-03-05', capabilities: { vision: false, reasoning: true, functionCalling: false }, @@ -2320,7 +2481,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://developers.cloudflare.com/workers-ai/models/', contextWindow: 262144, - maxOutputTokens: 262144, + maxOutputTokens: 256000, structuredOutput: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, @@ -2329,6 +2490,46 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, }, deepinfra: { + 'ByteDance/Seed-2.0-code': { + displayName: 'Seed 2.0 Code', + description: + 'ByteDance Seed coding model for multimodal software engineering and long-running agents', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 256000, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-02-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, + 'ByteDance/Seed-2.0-mini': { + displayName: 'Seed 2.0 Mini', + description: + 'Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 256000, + maxOutputTokens: 32000, + structuredOutput: true, + lastUpdated: '2026-02-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, + 'ByteDance/Seed-2.0-pro': { + displayName: 'Seed 2.0 Pro', + description: + 'Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 256000, + maxOutputTokens: 128000, + structuredOutput: true, + lastUpdated: '2026-02-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'deepseek-ai/DeepSeek-R1-0528': { displayName: 'DeepSeek-R1-0528', description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools', @@ -2354,6 +2555,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek-ai/DeepSeek-V3-0324': { + displayName: 'DeepSeek V3 0324', + description: 'March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 163840, + maxOutputTokens: 163840, + structuredOutput: true, + lastUpdated: '2025-03-24', + capabilities: { vision: false, reasoning: false, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'deepseek-ai/DeepSeek-V3.1': { displayName: 'DeepSeek-V3.1', description: 'Hybrid-reasoning DeepSeek model with thinking and non-thinking modes', @@ -2424,6 +2637,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek-ai/DeepSeek-V4-Pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 1048576, + maxOutputTokens: 384000, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'google/gemma-4-26B-A4B-it': { displayName: 'Gemma 4 26B A4B IT', description: 'Open Gemma instruction model for efficient chat and self-hosted deployments', @@ -2670,6 +2897,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'Qwen/Qwen3-30B-A3B': { + displayName: 'Qwen3 30B A3B', + description: + 'Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 40960, + maxOutputTokens: 16384, + structuredOutput: true, + lastUpdated: '2025-04-28', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'Qwen/Qwen3-32B': { displayName: 'Qwen3 32B', description: 'Dense open Qwen model for self-hosted chat, reasoning, and coding', @@ -2722,6 +2962,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'Qwen/Qwen3-VL-235B-A22B-Instruct': { + displayName: 'Qwen3 VL 235B A22B Instruct', + description: + 'Qwen vision-language instruct model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 262144, + maxOutputTokens: 32768, + knowledgeCutoff: '2025-03-31', + structuredOutput: true, + lastUpdated: '2025-09-23', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'Qwen/Qwen3.5-122B-A10B': { displayName: 'Qwen3.5 122B-A10B', description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', @@ -2821,6 +3075,34 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'Qwen/Qwen3.8-2.4T-A95B': { + displayName: 'Qwen3.8 2.4T A95B', + description: + 'Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 262144, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'Qwen/Qwen3.8-27B': { + displayName: 'Qwen3.8 27B', + description: + 'Dense 27B vision-language model for coding, agent tasks, and image and video understanding', + lifecycle: 'active', + docsUrl: 'https://deepinfra.com/models', + contextWindow: 262144, + maxOutputTokens: 32768, + structuredOutput: true, + lastUpdated: '2026-08-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'Qwen/Qwen3.8-Max': { displayName: 'Qwen3.8 Max', description: @@ -2878,6 +3160,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://deepinfra.com/models', contextWindow: 524288, maxOutputTokens: 1048576, + structuredOutput: true, lastUpdated: '2026-07-30', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'audio'], output: ['text'] }, @@ -3036,14 +3319,14 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'deepseek-v4-pro': { displayName: 'DeepSeek V4 Pro', - description: 'Open MoE flagship with million-token context for coding and long agent runs', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', lifecycle: 'active', docsUrl: 'https://api-docs.deepseek.com/quick_start/pricing', contextWindow: 1000000, maxOutputTokens: 384000, - knowledgeCutoff: '2025-05', structuredOutput: true, - lastUpdated: '2026-04-24', + lastUpdated: '2026-08-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, @@ -3093,6 +3376,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'accounts/fireworks/models/deepseek-v4-pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://fireworks.ai/docs/', + contextWindow: 1000000, + maxOutputTokens: 384000, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'accounts/fireworks/models/glm-5p2': { displayName: 'GLM 5.2', description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering', @@ -3131,6 +3428,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'accounts/fireworks/models/inkling': { + displayName: 'Inkling', + description: + 'Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio', + lifecycle: 'active', + docsUrl: 'https://fireworks.ai/docs/', + contextWindow: 1048576, + maxOutputTokens: 1048576, + lastUpdated: '2026-07-15', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image', 'audio'], output: ['text'] }, + }, 'accounts/fireworks/models/kimi-k2p6': { displayName: 'Kimi K2.6', description: 'Kimi reasoning model for long-horizon research, planning, and tool use', @@ -3193,32 +3502,86 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, - 'accounts/fireworks/models/qwen3p7-plus': { - displayName: 'Qwen 3.7 Plus', - description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + 'accounts/fireworks/models/muse-glimmer-30b': { + displayName: 'Muse Glimmer 30B', + description: + 'Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.', lifecycle: 'active', docsUrl: 'https://fireworks.ai/docs/', - contextWindow: 262144, - maxOutputTokens: 65536, - lastUpdated: '2026-06-12', + contextWindow: 131072, + maxOutputTokens: 131072, + knowledgeCutoff: '2026-01-04', + structuredOutput: true, + lastUpdated: '2026-08-10', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, - 'accounts/fireworks/routers/glm-5p2-fast': { - displayName: 'GLM 5.2 Fast', - description: 'Efficient GLM model for fast reasoning, coding, and agent workflows', + 'accounts/fireworks/models/nemotron-3-ultra-nvfp4': { + displayName: 'Nemotron 3 Ultra 550B A55B', + description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy', lifecycle: 'active', docsUrl: 'https://fireworks.ai/docs/', - contextWindow: 1048575, - maxOutputTokens: 131072, - lastUpdated: '2026-06-26', + contextWindow: 262144, + maxOutputTokens: 128000, + lastUpdated: '2026-06-04', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + thinkingOptions: { toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, - 'accounts/fireworks/routers/kimi-k2p6-fast': { - displayName: 'Kimi K2.6 Fast', + 'accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b': { + displayName: 'Nemotron 3.5 Lightning 30B A3B', + description: + 'Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads', + lifecycle: 'active', + docsUrl: 'https://fireworks.ai/docs/', + contextWindow: 262144, + maxOutputTokens: 262144, + structuredOutput: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'accounts/fireworks/models/qwen3p7-plus': { + displayName: 'Qwen 3.7 Plus', + description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: 'https://fireworks.ai/docs/', + contextWindow: 262144, + maxOutputTokens: 65536, + lastUpdated: '2026-06-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, + 'accounts/fireworks/models/qwen3p8-max': { + displayName: 'Qwen3.8 Max', + description: + '2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows', + lifecycle: 'active', + docsUrl: 'https://fireworks.ai/docs/', + contextWindow: 262144, + maxOutputTokens: 131072, + lastUpdated: '2026-08-03', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'accounts/fireworks/routers/glm-5p2-fast': { + displayName: 'GLM 5.2 Fast', + description: 'Efficient GLM model for fast reasoning, coding, and agent workflows', + lifecycle: 'active', + docsUrl: 'https://fireworks.ai/docs/', + contextWindow: 1048575, + maxOutputTokens: 131072, + lastUpdated: '2026-06-26', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'accounts/fireworks/routers/kimi-k2p6-fast': { + displayName: 'Kimi K2.6 Fast', description: 'Kimi reasoning model for long-horizon research, planning, and tool use', lifecycle: 'active', docsUrl: 'https://fireworks.ai/docs/', @@ -3465,6 +3828,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'gemini-3.7-flash': { + displayName: 'Gemini 3.7 Flash', + description: + 'High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning', + lifecycle: 'active', + docsUrl: 'https://docs.github.com/en/copilot', + contextWindow: 1000000, + inputLimit: 936000, + maxOutputTokens: 64000, + knowledgeCutoff: '2026-03', + structuredOutput: true, + lastUpdated: '2026-08-13', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'gpt-4.1': { displayName: 'GPT-4.1', description: @@ -3648,8 +4027,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'grok-4.5': { displayName: 'Grok 4.5', - description: - "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk", + description: "xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk", lifecycle: 'active', docsUrl: 'https://docs.github.com/en/copilot', contextWindow: 500000, @@ -3661,6 +4039,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'grok-4.6': { + displayName: 'Grok 4.6', + description: + "xAI's frontier model for long-running agents, coding, knowledge work, and visual projects", + lifecycle: 'active', + docsUrl: 'https://docs.github.com/en/copilot', + contextWindow: 500000, + inputLimit: 372000, + maxOutputTokens: 128000, + knowledgeCutoff: '2026-02-01', + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'kimi-k2.7-code': { displayName: 'Kimi K2.7 Code', description: @@ -3706,6 +4100,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'mai-code-1.1-flash': { + displayName: 'MAI-Code-1.1-Flash', + description: + 'Microsoft coding model with native vision support, optimized for fast and efficient software development', + lifecycle: 'active', + docsUrl: 'https://docs.github.com/en/copilot', + contextWindow: 256000, + inputLimit: 128000, + maxOutputTokens: 128000, + structuredOutput: true, + lastUpdated: '2026-08-11', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, }, google: { 'deep-research-max-preview-04-2026': { @@ -3733,32 +4142,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text', 'image'] }, }, - 'gemini-2.0-flash': { - displayName: 'Gemini 2.0 Flash', - description: 'Earlier Gemini Flash workhorse for responsive multimodal apps and tool use', - lifecycle: 'deprecated', - docsUrl: 'https://ai.google.dev/gemini-api/docs/models', - contextWindow: 1048576, - maxOutputTokens: 8192, - knowledgeCutoff: '2024-06', - structuredOutput: true, - lastUpdated: '2024-12-11', - capabilities: { vision: true, reasoning: false, functionCalling: true }, - modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, - }, - 'gemini-2.0-flash-lite': { - displayName: 'Gemini 2.0 Flash-Lite', - description: 'Legacy model retained for compatibility with older integrations', - lifecycle: 'deprecated', - docsUrl: 'https://ai.google.dev/gemini-api/docs/models', - contextWindow: 1048576, - maxOutputTokens: 8192, - knowledgeCutoff: '2024-06', - structuredOutput: true, - lastUpdated: '2024-12-11', - capabilities: { vision: true, reasoning: false, functionCalling: true }, - modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, - }, 'gemini-2.5-computer-use-preview-10-2025': { displayName: 'Gemini 2.5 Computer Use Preview 10-2025', description: 'Specialized Gemini 2.5 model for browser-control agents that automate UI tasks', @@ -3889,21 +4272,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text', 'image'] }, }, - 'gemini-3-pro-preview': { - displayName: 'Gemini 3 Pro Preview', - description: - 'Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts', - lifecycle: 'deprecated', - docsUrl: 'https://ai.google.dev/gemini-api/docs/models', - contextWindow: 1048576, - maxOutputTokens: 65536, - knowledgeCutoff: '2025-01', - structuredOutput: true, - lastUpdated: '2025-11-18', - capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'high'] }, - modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, - }, 'gemini-3.1-flash-image': { displayName: 'Nano Banana 2', description: 'Image model for prompt-driven generation, editing, and visual design workflows', @@ -4083,6 +4451,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, }, + 'gemini-3.7-flash': { + displayName: 'Gemini 3.7 Flash', + description: + 'High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning', + lifecycle: 'active', + docsUrl: 'https://ai.google.dev/gemini-api/docs/models', + contextWindow: 1048576, + maxOutputTokens: 65536, + knowledgeCutoff: '2026-03', + structuredOutput: true, + lastUpdated: '2026-08-13', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, + }, 'gemini-embedding-001': { displayName: 'Gemini Embedding 001', description: @@ -4199,6 +4582,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 65536, structuredOutput: false, + isFree: true, lastUpdated: '2026-03-25', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text', 'audio'] }, @@ -4212,6 +4596,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 65536, structuredOutput: false, + isFree: true, lastUpdated: '2026-03-25', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text', 'audio'] }, @@ -4258,6 +4643,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://console.groq.com/docs/models', contextWindow: 4096, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-01-23', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -4470,6 +4856,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek-ai/DeepSeek-V3-0324': { + displayName: 'DeepSeek V3 0324', + description: 'DeepSeek chat model for instruction following, coding, and analysis', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 163840, + maxOutputTokens: 163840, + structuredOutput: true, + lastUpdated: '2025-03-24', + capabilities: { vision: false, reasoning: false, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'deepseek-ai/DeepSeek-V3.1': { displayName: 'DeepSeek-V3.1', description: 'DeepSeek chat model for instruction following, coding, and analysis', @@ -4536,6 +4934,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek-ai/DeepSeek-V4-Pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: 'Flagship DeepSeek model for coding, reasoning, and agentic work', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 1000000, + maxOutputTokens: 384000, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'] }, + modalities: { input: ['text'], output: ['text'] }, + }, 'google/gemma-4-26B-A4B-it': { displayName: 'Gemma 4 26B A4B IT', description: 'Open Gemma instruction model for efficient chat and self-hosted deployments', @@ -4560,6 +4971,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'meta-llama/Llama-3.1-8B-Instruct': { + displayName: 'Llama-3.1-8B-Instruct', + description: 'Open Llama instruction model for multilingual chat, reasoning, and coding', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 131072, + maxOutputTokens: 4096, + knowledgeCutoff: '2023-12', + structuredOutput: true, + lastUpdated: '2024-07-23', + capabilities: { vision: false, reasoning: false, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'meta-llama/Llama-3.3-70B-Instruct': { displayName: 'Llama-3.3-70B-Instruct', description: 'Popular open Llama workhorse for multilingual chat, coding, and self-hosting', @@ -4746,6 +5170,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'Qwen/Qwen2.5-Coder-32B-Instruct': { + displayName: 'Qwen2.5-Coder-32B-Instruct', + description: 'Qwen coding model for software agents, repository edits, and code reasoning', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 131072, + maxOutputTokens: 8192, + structuredOutput: true, + lastUpdated: '2024-11-12', + capabilities: { vision: false, reasoning: false, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'Qwen/Qwen3-235B-A22B': { displayName: 'Qwen3 235B-A22B', description: 'Large open Qwen MoE for multilingual reasoning, coding, and tool use', @@ -4783,6 +5219,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'Qwen/Qwen3-30B-A3B': { + displayName: 'Qwen3 30B A3B', + description: 'Qwen instruction model for multilingual chat, reasoning, and tool use', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 40960, + maxOutputTokens: 16384, + structuredOutput: true, + lastUpdated: '2025-04-28', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'Qwen/Qwen3-32B': { displayName: 'Qwen3 32B', description: 'Dense open Qwen model for self-hosted chat, reasoning, and coding', @@ -4883,6 +5331,33 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'Qwen/Qwen3-VL-235B-A22B-Instruct': { + displayName: 'Qwen3 VL 235B A22B Instruct', + description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 131072, + maxOutputTokens: 32768, + knowledgeCutoff: '2025-03-31', + structuredOutput: true, + lastUpdated: '2025-09-23', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, + 'Qwen/Qwen3-VL-235B-A22B-Thinking': { + displayName: 'Qwen3 VL 235B A22B Thinking', + description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 131072, + maxOutputTokens: 32768, + knowledgeCutoff: '2025-03-31', + structuredOutput: true, + lastUpdated: '2025-09-23', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'Qwen/Qwen3.5-122B-A10B': { displayName: 'Qwen3.5 122B-A10B', description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', @@ -4968,6 +5443,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'Qwen/Qwen3.8-2.4T-A95B': { + displayName: 'Qwen3.8 2.4T A95B', + description: 'Qwen instruction model for multilingual chat, reasoning, and tool use', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 262144, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] }, + modalities: { input: ['text'], output: ['text'] }, + }, 'stepfun-ai/Step-3.5-Flash': { displayName: 'Step 3.5 Flash', description: 'StepFun flash lane for quick multimodal reasoning and coding assistance', @@ -5118,6 +5606,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'zai-org/GLM-4.6V-Flash': { + displayName: 'GLM-4.6V-Flash', + description: 'GLM vision model for visual reasoning, documents, and multimodal agents', + lifecycle: 'active', + docsUrl: 'https://huggingface.co/docs/inference-providers', + contextWindow: 131072, + maxOutputTokens: 32768, + lastUpdated: '2025-12-08', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'zai-org/GLM-4.7': { displayName: 'GLM-4.7', description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering', @@ -5138,6 +5638,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 128000, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-08-08', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5188,6 +5689,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-07-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, @@ -5202,6 +5704,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-07-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'high', 'max'] }, @@ -5217,6 +5720,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 32768, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-12', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -5230,6 +5734,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 32768, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-12', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -5405,6 +5910,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 196608, maxOutputTokens: 128000, + isFree: true, lastUpdated: '2025-10-27', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5416,6 +5922,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 204800, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2025-12-23', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5427,6 +5934,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 204800, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5438,6 +5946,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 204800, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2026-02-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5449,6 +5958,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 204800, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5460,6 +5970,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 204800, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -5471,6 +5982,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.minimax.io/docs/token-plan/intro', contextWindow: 1000000, maxOutputTokens: 128000, + isFree: true, lastUpdated: '2026-06-25', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -5571,6 +6083,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 256000, knowledgeCutoff: '2025-12', + isFree: true, lastUpdated: '2025-12-09', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6022,6 +6535,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-05-22', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6033,6 +6547,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 8192, maxOutputTokens: 1024, + isFree: true, lastUpdated: '2026-04-30', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6044,6 +6559,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 40960, maxOutputTokens: 40960, + isFree: true, lastUpdated: '2025-08-12', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['image'] }, @@ -6058,6 +6574,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 0, knowledgeCutoff: '2024-07', structuredOutput: false, + isFree: true, lastUpdated: '2026-02-04', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -6070,6 +6587,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 40960, maxOutputTokens: 40960, knowledgeCutoff: '2025-06', + isFree: true, lastUpdated: '2026-01-31', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['image', 'text'], output: ['image'] }, @@ -6082,6 +6600,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 4096, maxOutputTokens: 0, knowledgeCutoff: '2024-08', + isFree: true, lastUpdated: '2025-09-05', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['image'] }, @@ -6095,6 +6614,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262000, maxOutputTokens: 262000, structuredOutput: true, + isFree: true, lastUpdated: '2025-11-25', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6135,6 +6655,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, structuredOutput: true, + isFree: true, lastUpdated: '2024-07-16', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6146,6 +6667,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-03-12', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6157,6 +6679,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-03-12', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6170,6 +6693,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 4096, knowledgeCutoff: '2024-06', structuredOutput: true, + isFree: true, lastUpdated: '2025-06-12', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6183,6 +6707,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 4096, knowledgeCutoff: '2024-06', structuredOutput: true, + isFree: true, lastUpdated: '2025-06-03', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6195,6 +6720,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 16384, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6207,6 +6733,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2024-08-26', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6218,6 +6745,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-03-10', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6229,6 +6757,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-06-12', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6241,6 +6770,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, structuredOutput: true, + isFree: true, lastUpdated: '2024-07-16', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6253,6 +6783,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 16000, maxOutputTokens: 4096, knowledgeCutoff: '2023-12', + isFree: true, lastUpdated: '2025-01-01', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6266,6 +6797,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 4096, knowledgeCutoff: '2023-12', structuredOutput: true, + isFree: true, lastUpdated: '2024-09-18', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6279,6 +6811,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 4096, knowledgeCutoff: '2023-12', structuredOutput: true, + isFree: true, lastUpdated: '2024-09-18', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6291,6 +6824,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 32768, maxOutputTokens: 32000, structuredOutput: true, + isFree: true, lastUpdated: '2024-09-18', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6303,6 +6837,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 8192, knowledgeCutoff: '2023-12', + isFree: true, lastUpdated: '2024-09-25', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6315,6 +6850,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, structuredOutput: true, + isFree: true, lastUpdated: '2024-11-26', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6328,6 +6864,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 4096, knowledgeCutoff: '2024-02', structuredOutput: true, + isFree: true, lastUpdated: '2025-04-01', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6340,10 +6877,27 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-04-30', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'meta/muse-glimmer-30b': { + displayName: 'Muse Glimmer 30B', + description: + 'Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.', + lifecycle: 'active', + docsUrl: 'https://docs.api.nvidia.com/nim/', + contextWindow: 131072, + maxOutputTokens: 131072, + knowledgeCutoff: '2026-01-04', + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-10', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'max'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'microsoft/phi-4-mini-instruct': { displayName: 'Phi-4-Mini', description: 'Efficient model for low-latency assistance, extraction, and routine automation', @@ -6352,6 +6906,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 8192, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2025-09-05', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6365,6 +6920,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< inputLimit: 128000, maxOutputTokens: 16384, structuredOutput: false, + isFree: true, lastUpdated: '2025-07-26', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6376,6 +6932,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 204800, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2026-04-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6388,6 +6945,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 1000000, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-06-01', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6402,6 +6960,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< inputLimit: 32768, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2025-09-25', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6413,6 +6972,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 262144, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-12-02', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6425,6 +6985,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 65536, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2025-04-01', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6438,6 +6999,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2025-12-02', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6451,6 +7013,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< inputLimit: 131072, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2025-09-25', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6463,6 +7026,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-29', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['none', 'high'] }, @@ -6475,6 +7039,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-06-12', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6487,6 +7052,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 8192, structuredOutput: true, + isFree: true, lastUpdated: '2026-03-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['none', 'high'] }, @@ -6499,6 +7065,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 65536, maxOutputTokens: 13108, + isFree: true, lastUpdated: '2024-04-17', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6510,6 +7077,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 32768, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-03-15', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6522,6 +7090,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 262144, knowledgeCutoff: '2024-10', + isFree: true, lastUpdated: '2025-09-05', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6535,6 +7104,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-04-21', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'xhigh', 'max'] }, @@ -6547,6 +7117,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-04-16', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: [], output: ['text'] }, @@ -6558,6 +7129,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-07-20', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: [], output: ['text'] }, @@ -6569,6 +7141,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-03-18', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: [] }, @@ -6581,6 +7154,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-12-01', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6592,6 +7166,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-06-30', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: [] }, @@ -6603,6 +7178,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-02-26', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: [] }, @@ -6614,6 +7190,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-03-03', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6626,6 +7203,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 32768, maxOutputTokens: 2048, + isFree: true, lastUpdated: '2025-07-24', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6637,6 +7215,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-04-15', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6648,6 +7227,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6660,6 +7240,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 32768, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-04-10', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6672,6 +7253,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-10-28', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6683,6 +7265,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2025-04-07', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6695,6 +7278,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 131072, maxOutputTokens: 65536, + isFree: true, lastUpdated: '2025-04-07', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6707,6 +7291,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 131072, maxOutputTokens: 65536, + isFree: true, lastUpdated: '2025-07-25', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6720,6 +7305,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 32768, maxOutputTokens: 2048, + isFree: true, lastUpdated: '2026-02-10', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6732,6 +7318,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-03-31', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6743,6 +7330,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-06-12', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'audio'], output: ['audio'] }, @@ -6755,6 +7343,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-04-16', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6767,6 +7356,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 131072, knowledgeCutoff: '2024-09', + isFree: true, lastUpdated: '2024-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6780,6 +7370,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6812,6 +7403,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'nvidia/nemotron-3.5-lightning-30b-a3b': { + displayName: 'Nemotron 3.5 Lightning 30B A3B', + description: + 'Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads', + lifecycle: 'active', + docsUrl: 'https://docs.api.nvidia.com/nim/', + contextWindow: 262144, + maxOutputTokens: 262144, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'nvidia/nemotron-content-safety-reasoning-4b': { displayName: 'nemotron-content-safety-reasoning-4b', description: @@ -6820,6 +7426,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-01-22', capabilities: { vision: false, reasoning: true, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6831,6 +7438,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2024-08-26', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -6842,6 +7450,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 128000, + isFree: true, lastUpdated: '2025-10-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -6853,6 +7462,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-03-16', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'audio'], output: ['text'] }, @@ -6865,6 +7475,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 32768, maxOutputTokens: 2048, + isFree: true, lastUpdated: '2025-07-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6876,6 +7487,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 32768, maxOutputTokens: 2048, + isFree: true, lastUpdated: '2025-05-29', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6888,6 +7500,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 131072, knowledgeCutoff: '2024-09', + isFree: true, lastUpdated: '2025-08-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -6901,6 +7514,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-01-17', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6913,6 +7527,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-12-12', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6924,6 +7539,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-07-20', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: [], output: ['text'] }, @@ -6935,6 +7551,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-11-13', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: [], output: ['text'] }, @@ -6946,6 +7563,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-06-13', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6957,6 +7575,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-04-16', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: [], output: ['text'] }, @@ -6968,6 +7587,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2026-01-01', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6979,6 +7599,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 0, maxOutputTokens: 4096, + isFree: true, lastUpdated: '2025-01-08', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text'] }, @@ -6992,6 +7613,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 8192, knowledgeCutoff: '2025-08', structuredOutput: true, + isFree: true, lastUpdated: '2025-08-14', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -7006,6 +7628,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 32768, structuredOutput: true, + isFree: true, lastUpdated: '2025-08-05', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -7019,6 +7642,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 0, maxOutputTokens: 4096, knowledgeCutoff: '2023-09', + isFree: true, lastUpdated: '2025-09-05', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['audio'], output: ['text'] }, @@ -7031,6 +7655,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 16384, structuredOutput: false, + isFree: true, lastUpdated: '2026-07-02', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -7043,6 +7668,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 0, maxOutputTokens: 0, structuredOutput: false, + isFree: true, lastUpdated: '2025-08-07', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['image'] }, @@ -7055,6 +7681,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 0, maxOutputTokens: 0, structuredOutput: false, + isFree: true, lastUpdated: '2025-08-19', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['image'] }, @@ -7067,6 +7694,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, structuredOutput: true, + isFree: true, lastUpdated: '2024-11-06', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -7079,6 +7707,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 66536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-07-23', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -7091,6 +7720,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 16384, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2025-09-05', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -7103,6 +7733,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 65536, structuredOutput: true, + isFree: true, lastUpdated: '2026-02-23', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -7117,6 +7748,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 8192, knowledgeCutoff: '2026-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-02-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -7130,6 +7762,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-07-25', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -7141,6 +7774,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 256000, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-02-02', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -7153,6 +7787,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 256000, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-05-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh', 'max'] }, @@ -7166,6 +7801,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 1048576, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-07-15', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'audio'], output: ['text'] }, @@ -7178,6 +7814,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://docs.api.nvidia.com/nim/', contextWindow: 128000, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2025-04-10', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -7191,6 +7828,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -8130,6 +8768,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 32000, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2025-10-17', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -8301,7 +8940,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'deepseek-v4-flash': { - displayName: 'DeepSeek V4 Flash (New)', + displayName: 'DeepSeek V4 Flash', description: 'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding', lifecycle: 'active', @@ -8316,7 +8955,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text'], output: ['text'] }, }, 'deepseek-v4-flash-free': { - displayName: 'DeepSeek V4 Flash Free (New)', + displayName: 'DeepSeek V4 Flash Free', description: 'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding', lifecycle: 'active', @@ -8325,6 +8964,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 128000, knowledgeCutoff: '2025-05', structuredOutput: true, + isFree: true, lastUpdated: '2026-07-31', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'high', 'max'] }, @@ -8428,15 +9068,30 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, }, - 'glm-4.6': { - displayName: 'GLM-4.6', - description: 'Legacy model retained for compatibility with older integrations', - lifecycle: 'deprecated', + 'gemini-3.7-flash': { + displayName: 'Gemini 3.7 Flash', + description: + 'High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning', + lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', - contextWindow: 204800, - maxOutputTokens: 131072, - knowledgeCutoff: '2025-04', - lastUpdated: '2025-09-30', + contextWindow: 1048576, + maxOutputTokens: 65536, + knowledgeCutoff: '2026-03', + structuredOutput: true, + lastUpdated: '2026-08-13', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, + }, + 'glm-4.6': { + displayName: 'GLM-4.6', + description: 'Legacy model retained for compatibility with older integrations', + lifecycle: 'deprecated', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 204800, + maxOutputTokens: 131072, + knowledgeCutoff: '2025-04', + lastUpdated: '2025-09-30', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, modalities: { input: ['text'], output: ['text'] }, @@ -8462,6 +9117,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 204800, maxOutputTokens: 131072, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-12-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -8488,6 +9144,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 204800, maxOutputTokens: 131072, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-02-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -8798,7 +9455,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'gpt-5.6-sol': { - displayName: 'GPT-5.6 Sol', + displayName: 'GPT-5.6 Sol (50% Off)', description: 'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows', lifecycle: 'active', @@ -8830,8 +9487,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'grok-4.5': { displayName: 'Grok 4.5', - description: - "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk", + description: "xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk", lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 500000, @@ -8842,17 +9498,32 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'grok-4.6': { + displayName: 'Grok 4.6', + description: + "xAI's frontier model for long-running agents, coding, knowledge work, and visual projects", + lifecycle: 'active', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 500000, + maxOutputTokens: 500000, + knowledgeCutoff: '2026-02-01', + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'grok-build-0.1': { displayName: 'Grok Build 0.1', - description: 'Grok coding model for agentic engineering, edits, and codebase workflows', + description: 'Fast Grok coding model tuned for agentic engineering and iterative edits', lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 256000, maxOutputTokens: 256000, structuredOutput: true, - lastUpdated: '2026-05-20', + lastUpdated: '2026-04-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, - modalities: { input: ['text', 'image'], output: ['text'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'grok-code': { displayName: 'Grok Code Fast 1', @@ -8861,6 +9532,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 256000, maxOutputTokens: 256000, + isFree: true, lastUpdated: '2025-08-20', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -8868,12 +9540,15 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'hy3-free': { displayName: 'Hy3 Free', description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks', - lifecycle: 'deprecated', + lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 190000, maxOutputTokens: 64000, - lastUpdated: '2026-06-26', + structuredOutput: true, + isFree: true, + lastUpdated: '2026-07-06', capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'hy3-preview-free': { @@ -8884,6 +9559,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 64000, knowledgeCutoff: '2025-06', + isFree: true, lastUpdated: '2026-04-20', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -8933,6 +9609,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 262144, knowledgeCutoff: '2024-10', + isFree: true, lastUpdated: '2026-01-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -8982,11 +9659,12 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'laguna-s-2.1-free': { displayName: 'Laguna S 2.1 Free', description: 'Agentic coding model from Poolside in the XS size class for local deployment', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 256000, maxOutputTokens: 32000, structuredOutput: false, + isFree: true, lastUpdated: '2026-07-21', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -9000,6 +9678,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262100, maxOutputTokens: 32800, knowledgeCutoff: '2025-06', + isFree: true, lastUpdated: '2026-04-21', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9007,24 +9686,40 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'ling-3.0-flash-free': { displayName: 'Ling-3.0-flash Free', description: 'Efficient model for low-latency assistance, extraction, and routine automation', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 262144, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2026-07-23', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'ling-3.0-tiny-free': { + displayName: 'Ling-3.0-tiny Free', + description: + 'Compact MoE model for responsive agents, instruction following, and multi-turn conversations', + lifecycle: 'deprecated', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 262144, + maxOutputTokens: 32768, + structuredOutput: false, + isFree: true, + lastUpdated: '2026-08-06', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'longcat-2.0-free': { displayName: 'LongCat-2.0 Free', description: 'Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 1000000, maxOutputTokens: 131072, + isFree: true, lastUpdated: '2026-06-30', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -9038,6 +9733,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 65536, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2025-12-16', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9050,6 +9746,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 64000, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, @@ -9062,6 +9759,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 64000, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9074,6 +9772,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 32000, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-24', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'audio'], output: ['text'] }, @@ -9098,6 +9797,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 204800, maxOutputTokens: 131072, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2025-12-23', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9122,6 +9822,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 204800, maxOutputTokens: 131072, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9158,11 +9859,41 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 32000, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2026-05-31', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'muse-spark-1.2': { + displayName: 'Muse Spark 1.2', + description: + 'Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.', + lifecycle: 'active', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 1048576, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-05', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] }, + }, + 'muse-spark-1.2-contributor-free': { + displayName: 'Muse Spark 1.2 Free', + description: + 'Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.', + lifecycle: 'active', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 1048576, + maxOutputTokens: 131072, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-05', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] }, + }, 'nemotron-3-super-free': { displayName: 'Nemotron 3 Super Free', description: @@ -9172,6 +9903,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 204800, maxOutputTokens: 128000, knowledgeCutoff: '2026-02', + isFree: true, lastUpdated: '2026-03-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9184,19 +9916,35 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 128000, knowledgeCutoff: '2026-02', + isFree: true, lastUpdated: '2026-06-04', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'nemotron-3.5-lightning-free': { + displayName: 'Nemotron 3.5 Lightning Free', + description: + 'Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads', + lifecycle: 'active', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 262144, + maxOutputTokens: 262144, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'north-mini-code-free': { displayName: 'North Mini Code Free', description: 'Cohere coding model for practical software engineering and agentic edits', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 256000, maxOutputTokens: 64000, knowledgeCutoff: '2025-09-23', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-09', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['none', 'high'] }, @@ -9248,6 +9996,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 65536, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -9261,6 +10010,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262000, maxOutputTokens: 66000, knowledgeCutoff: '2025-06', + isFree: true, lastUpdated: '2026-05-08', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9273,6 +10023,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 131072, knowledgeCutoff: '2025-06', + isFree: true, lastUpdated: '2026-01-28', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -9280,7 +10031,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'opencode-go': { 'deepseek-v4-flash': { - displayName: 'DeepSeek V4 Flash (New)', + displayName: 'DeepSeek V4 Flash', description: 'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding', lifecycle: 'active', @@ -9295,7 +10046,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text'], output: ['text'] }, }, 'deepseek-v4-pro': { - displayName: 'DeepSeek V4 Pro', + displayName: 'DeepSeek V4 Pro (New)', description: 'Flagship DeepSeek model for coding, reasoning, and agentic work', lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', @@ -9346,8 +10097,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high', 'max'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'glm-5.3': { + displayName: 'GLM-5.3', + description: + 'Flagship GLM model for long-horizon coding, agents, and complex project delivery', + lifecycle: 'active', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 1000000, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-14', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'] }, + modalities: { input: ['text'], output: ['text'] }, + }, 'gpt-5.6-luna': { - displayName: 'GPT-5.6 Luna (2x usage)', + displayName: 'GPT-5.6 Luna', description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads', lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', @@ -9363,8 +10128,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'grok-4.5': { displayName: 'Grok 4.5', - description: - "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk", + description: "xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk", lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', contextWindow: 500000, @@ -9376,7 +10140,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text', 'image'], output: ['text'] }, }, hy3: { - displayName: 'Hy3', + displayName: 'Hy3 (8x usage)', description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks', lifecycle: 'active', docsUrl: 'https://opencode.ai/docs/zen', @@ -9524,6 +10288,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'muse-spark-1.2-contributor': { + displayName: 'Muse Spark 1.2 Contributor', + description: + 'Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.', + lifecycle: 'active', + docsUrl: 'https://opencode.ai/docs/zen', + contextWindow: 1048576, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-05', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] }, + }, 'qwen3.5-plus': { displayName: 'Qwen3.5 Plus', description: 'Legacy model retained for compatibility with older integrations', @@ -9627,7 +10405,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-04-21', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, '~anthropic/claude-sonnet-latest': { @@ -9641,7 +10419,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-04-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, '~deepseek/deepseek-v4-flash-latest': { @@ -9649,12 +10427,12 @@ export const GENERATED_MODELS_DEV_METADATA: Record< description: 'Fast DeepSeek model for efficient chat, coding help, and agent loops', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 1048576, - maxOutputTokens: 65536, + contextWindow: 1310720, + maxOutputTokens: 262144, structuredOutput: true, lastUpdated: '2026-08-01', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'high', 'max'] }, + thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, '~google/gemini-flash-latest': { @@ -9668,7 +10446,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-04-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] }, }, '~google/gemini-pro-latest': { @@ -9691,11 +10469,11 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 1048576, - maxOutputTokens: 1048576, + maxOutputTokens: 974842, structuredOutput: true, lastUpdated: '2026-04-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'high', 'max'] }, + thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, '~openai/gpt-latest': { @@ -9736,20 +10514,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-07-08', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, - 'ai21/jamba-large-1.7': { - displayName: 'Jamba Large 1.7', - description: 'Flagship model for demanding analysis, coding, and production agent workflows', + '~z-ai/glm-latest': { + displayName: 'GLM Latest', + description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 256000, - maxOutputTokens: 4096, - knowledgeCutoff: '2024-08-31', + contextWindow: 1048576, + maxOutputTokens: 131072, structuredOutput: false, - lastUpdated: '2025-08-08', - capabilities: { vision: false, reasoning: false, functionCalling: true }, + lastUpdated: '2026-08-19', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'] }, modalities: { input: ['text'], output: ['text'] }, }, 'aion-labs/aion-2.0': { @@ -9885,8 +10663,8 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'Open-weight instruction model for adaptable chat and self-hosted production workloads', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 16384, - maxOutputTokens: 2048, + contextWindow: 32768, + maxOutputTokens: 4096, knowledgeCutoff: '2024-06-30', structuredOutput: true, lastUpdated: '2024-10-22', @@ -9988,7 +10766,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-03-13', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-opus-4.7': { @@ -10002,7 +10780,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-04-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-opus-4.7-fast': { @@ -10016,7 +10794,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-04-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-opus-4.8': { @@ -10031,7 +10809,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-05-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-opus-4.8-fast': { @@ -10045,7 +10823,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-05-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-opus-5': { @@ -10059,7 +10837,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-07-24', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-opus-5-fast': { @@ -10073,7 +10851,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-07-24', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-sonnet-4': { @@ -10116,7 +10894,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-03-13', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'anthropic/claude-sonnet-5': { @@ -10130,18 +10908,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-06-30', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, 'arcee-ai/trinity-large-thinking': { displayName: 'Trinity Large Thinking', - description: 'Flagship model for demanding analysis, coding, and production agent workflows', + description: + 'Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, maxOutputTokens: 262144, structuredOutput: true, - lastUpdated: '2026-04-01', + lastUpdated: '2026-05-28', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, @@ -10195,30 +10974,56 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['image', 'text'], output: ['text'] }, }, + 'bytedance-seed/seed-2-1-turbo': { + displayName: 'Seed 2.1 Turbo', + description: 'Multimodal reasoning model for visual analysis, planning, and tool use', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 262144, + maxOutputTokens: 262144, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, + 'bytedance-seed/seed-2.0-code': { + displayName: 'Seed 2.0 Code', + description: + 'Coding model for repository understanding, refactors, and agentic engineering tasks', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 262144, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-02-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'bytedance-seed/seed-2.0-lite': { - displayName: 'Seed-2.0-Lite', + displayName: 'Seed 2.0 Lite', description: 'Multimodal reasoning model for visual analysis, planning, and tool use', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, maxOutputTokens: 131072, structuredOutput: true, - lastUpdated: '2026-03-10', + lastUpdated: '2026-02-14', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'], toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, 'bytedance-seed/seed-2.0-mini': { - displayName: 'Seed-2.0-Mini', + displayName: 'Seed 2.0 Mini', description: 'Multimodal reasoning model for visual analysis, planning, and tool use', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, maxOutputTokens: 131072, structuredOutput: true, - lastUpdated: '2026-02-26', + lastUpdated: '2026-02-14', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'], toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, 'bytedance/ui-tars-1.5-7b': { @@ -10307,6 +11112,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 64000, structuredOutput: false, + isFree: true, lastUpdated: '2026-06-17', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -10343,7 +11149,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 163840, - maxOutputTokens: 65536, + maxOutputTokens: 163840, knowledgeCutoff: '2024-07-31', structuredOutput: true, lastUpdated: '2025-03-24', @@ -10370,7 +11176,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'Classic open reasoning model for transparent math, coding, and deliberate problem solving', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 163840, + contextWindow: 64000, maxOutputTokens: 16000, knowledgeCutoff: '2024-07', structuredOutput: true, @@ -10410,7 +11216,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 163840, - maxOutputTokens: 32768, + maxOutputTokens: 163840, knowledgeCutoff: '2025-03-31', structuredOutput: true, lastUpdated: '2025-09-22', @@ -10452,12 +11258,12 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 1048576, - maxOutputTokens: 131072, + maxOutputTokens: 384000, knowledgeCutoff: '2025-05', structuredOutput: true, lastUpdated: '2026-04-24', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['high', 'xhigh'] }, + thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'deepseek/deepseek-v4-flash-0731': { @@ -10466,13 +11272,13 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 1048576, - maxOutputTokens: 65536, + contextWindow: 1310720, + maxOutputTokens: 393216, knowledgeCutoff: '2025-05', structuredOutput: true, lastUpdated: '2026-07-31', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'high', 'max'] }, + thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'deepseek/deepseek-v4-pro': { @@ -10481,14 +11287,41 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 1048576, - maxOutputTokens: 384000, + maxOutputTokens: 393216, knowledgeCutoff: '2025-05', structuredOutput: true, lastUpdated: '2026-04-24', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['high', 'xhigh'] }, + thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'deepseek/deepseek-v4-pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 1048576, + maxOutputTokens: 384000, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'dots-studio/dots-3-note-preview:free': { + displayName: 'Dots3-Note Preview (free)', + description: 'Multimodal reasoning model for visual analysis, planning, and tool use', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 512000, + maxOutputTokens: 512000, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'google/gemini-2.5-flash': { displayName: 'Gemini 2.5 Flash', description: 'Fast Gemini workhorse for multimodal apps where latency and price matter', @@ -10582,7 +11415,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2025-12-17', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'], toggle: true }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, }, 'google/gemini-3-pro-image': { @@ -10622,8 +11455,8 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-05-28', capabilities: { vision: true, reasoning: true, functionCalling: false }, - thinkingOptions: { efforts: ['minimal', 'high'] }, - modalities: { input: ['text', 'image'], output: ['text', 'image'] }, + thinkingOptions: { efforts: ['minimal', 'high'], toggle: true }, + modalities: { input: ['image', 'text'], output: ['text', 'image'] }, }, 'google/gemini-3.1-flash-image-preview': { displayName: 'Nano Banana 2', @@ -10636,7 +11469,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-02-26', capabilities: { vision: true, reasoning: true, functionCalling: false }, - thinkingOptions: { efforts: ['minimal', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'high'], toggle: true }, modalities: { input: ['image', 'text'], output: ['text', 'image'] }, }, 'google/gemini-3.1-flash-lite': { @@ -10650,7 +11483,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-05-07', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'], toggle: true }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, }, 'google/gemini-3.1-flash-lite-image': { @@ -10664,7 +11497,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: false, lastUpdated: '2026-06-30', capabilities: { vision: true, reasoning: true, functionCalling: false }, - thinkingOptions: { efforts: ['minimal', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'high'], toggle: true }, modalities: { input: ['text', 'image'], output: ['text', 'image'] }, }, 'google/gemini-3.1-flash-lite-preview': { @@ -10678,7 +11511,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-03-03', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'], toggle: true }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, }, 'google/gemini-3.1-pro-preview': { @@ -10751,12 +11584,26 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, }, - 'google/gemma-2-27b-it': { - displayName: 'Gemma 2 27B', - description: 'Open Gemma instruction model for efficient chat and self-hosted deployments', + 'google/gemini-3.7-flash': { + displayName: 'Gemini 3.7 Flash', + description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 8192, + contextWindow: 1048576, + maxOutputTokens: 65536, + knowledgeCutoff: '2026-03', + structuredOutput: true, + lastUpdated: '2026-08-13', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] }, + }, + 'google/gemma-2-27b-it': { + displayName: 'Gemma 2 27B', + description: 'Open Gemma instruction model for efficient chat and self-hosted deployments', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 8192, maxOutputTokens: 2048, knowledgeCutoff: '2024-06-30', structuredOutput: true, @@ -10837,6 +11684,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, structuredOutput: true, + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -10848,7 +11696,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, - maxOutputTokens: 262144, + maxOutputTokens: 16384, structuredOutput: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, @@ -10863,6 +11711,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2026-04-02', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -10876,6 +11725,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 65536, structuredOutput: false, + isFree: true, lastUpdated: '2026-03-25', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text', 'audio'] }, @@ -10888,6 +11738,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 65536, structuredOutput: false, + isFree: true, lastUpdated: '2026-03-25', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text', 'audio'] }, @@ -10982,18 +11833,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, - 'inclusionai/ling-3.0-tiny:free': { - displayName: 'Ling 3.0 Tiny (free)', - description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads', - lifecycle: 'active', - docsUrl: 'https://openrouter.ai/models', - contextWindow: 262144, - maxOutputTokens: 32768, - structuredOutput: false, - lastUpdated: '2026-08-06', - capabilities: { vision: false, reasoning: true, functionCalling: true }, - modalities: { input: ['text'], output: ['text'] }, - }, 'inclusionai/ring-2.6-1t': { displayName: 'Ring-2.6-1T', description: @@ -11047,17 +11886,17 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, - 'mancer/weaver': { - displayName: 'Weaver (alpha)', - description: 'General-purpose chat model for instruction following, writing, and analysis', + 'liquid/lfm-2.5-2.6b:free': { + displayName: 'LFM2.5-2.6B (free)', + description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', - contextWindow: 8000, - maxOutputTokens: 2000, - knowledgeCutoff: '2023-06-30', + contextWindow: 128000, + maxOutputTokens: 8192, structuredOutput: true, - lastUpdated: '2023-08-02', - capabilities: { vision: false, reasoning: false, functionCalling: false }, + isFree: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, 'meituan/longcat-2.0': { @@ -11087,13 +11926,13 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text'], output: ['text'] }, }, 'meta-llama/llama-3.1-8b-instruct': { - displayName: 'Llama 3.1 8B Instruct', + displayName: 'Llama-3.1-8B-Instruct', description: 'Open Llama instruction model for multilingual chat, reasoning, and coding', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 131072, maxOutputTokens: 131072, - knowledgeCutoff: '2023-12-31', + knowledgeCutoff: '2023-12', structuredOutput: true, lastUpdated: '2024-07-23', capabilities: { vision: false, reasoning: false, functionCalling: true }, @@ -11178,6 +12017,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['image', 'text'], output: ['text'] }, }, + 'meta/muse-glimmer-30b': { + displayName: 'Muse Glimmer 30B', + description: + 'Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 131072, + maxOutputTokens: 131072, + knowledgeCutoff: '2026-01-04', + structuredOutput: true, + lastUpdated: '2026-08-10', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'meta/muse-spark-1.1': { displayName: 'Muse Spark 1.1', description: 'Open Llama multimodal model for image understanding and text reasoning', @@ -11272,7 +12126,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< modalities: { input: ['text'], output: ['text'] }, }, 'minimax/minimax-m2-her': { - displayName: 'MiniMax M2-her', + displayName: 'MiniMax-M2 Her', description: 'MiniMax model for chat, coding, office work, and agentic tasks', lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', @@ -11659,7 +12513,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-07-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'high', 'max'] }, + thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, 'morph/morph-v3-fast': { @@ -11786,6 +12640,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 256000, structuredOutput: false, + isFree: true, lastUpdated: '2025-12-15', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -11798,6 +12653,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 65536, structuredOutput: false, + isFree: true, lastUpdated: '2026-04-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'audio'], output: ['text'] }, @@ -11813,7 +12669,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-03-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium'] }, + thinkingOptions: { efforts: ['low', 'medium'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'nvidia/nemotron-3-super-120b-a12b:free': { @@ -11825,9 +12681,10 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 262144, structuredOutput: true, + isFree: true, lastUpdated: '2026-03-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium'] }, + thinkingOptions: { efforts: ['low', 'medium'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'nvidia/nemotron-3-ultra-550b-a55b': { @@ -11840,7 +12697,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-06-04', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['medium', 'high'] }, + thinkingOptions: { efforts: ['medium', 'high'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'nvidia/nemotron-3-ultra-550b-a55b:free': { @@ -11851,9 +12708,10 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 65536, structuredOutput: false, + isFree: true, lastUpdated: '2026-06-04', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['medium', 'high'] }, + thinkingOptions: { efforts: ['medium', 'high'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, 'nvidia/nemotron-3.5-content-safety:free': { @@ -11865,10 +12723,38 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 8192, structuredOutput: false, + isFree: true, lastUpdated: '2026-06-04', capabilities: { vision: true, reasoning: true, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'nvidia/nemotron-3.5-lightning': { + displayName: 'Nemotron 3.5 Lightning 30B A3B', + description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 1000000, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'nvidia/nemotron-3.5-lightning:free': { + displayName: 'Nemotron 3.5 Lightning (free)', + description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 1000000, + maxOutputTokens: 65536, + structuredOutput: false, + isFree: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'nvidia/nemotron-nano-12b-v2-vl:free': { displayName: 'Nemotron Nano 12B 2 VL (free)', description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows', @@ -11877,6 +12763,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 128000, structuredOutput: false, + isFree: true, lastUpdated: '2025-10-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -11889,6 +12776,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 128000, structuredOutput: true, + isFree: true, lastUpdated: '2025-08-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -12246,7 +13134,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2025-11-13', capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, 'openai/gpt-5.2': { @@ -12272,7 +13160,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 128000, - maxOutputTokens: 16384, + maxOutputTokens: 32000, knowledgeCutoff: '2025-08-31', structuredOutput: true, lastUpdated: '2025-12-10', @@ -12310,19 +13198,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] }, modalities: { input: ['image', 'text', 'pdf'], output: ['text'] }, }, - 'openai/gpt-5.3-chat': { - displayName: 'GPT-5.3 Chat', - description: - 'Chat-tuned GPT model for conversational assistance, writing, and tool workflows', - lifecycle: 'active', - docsUrl: 'https://openrouter.ai/models', - contextWindow: 128000, - maxOutputTokens: 16384, - structuredOutput: true, - lastUpdated: '2026-03-03', - capabilities: { vision: true, reasoning: false, functionCalling: true }, - modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, - }, 'openai/gpt-5.3-codex': { displayName: 'GPT-5.3 Codex', description: @@ -12605,6 +13480,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 32768, structuredOutput: true, + isFree: true, lastUpdated: '2025-08-05', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -12769,6 +13645,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< inputLimit: 200000, maxOutputTokens: 8000, structuredOutput: true, + isFree: true, lastUpdated: '2026-02-01', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -12891,6 +13768,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2026-07-21', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -12915,6 +13793,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2026-07-02', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -13016,7 +13895,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 131072, - maxOutputTokens: 8192, + maxOutputTokens: 16384, knowledgeCutoff: '2025-03-31', structuredOutput: true, lastUpdated: '2025-04-28', @@ -13070,8 +13949,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 131072, - maxOutputTokens: 16384, - knowledgeCutoff: '2025-03-31', + maxOutputTokens: 8192, structuredOutput: false, lastUpdated: '2025-04-28', capabilities: { vision: false, reasoning: true, functionCalling: true }, @@ -13151,7 +14029,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, - maxOutputTokens: 32768, + maxOutputTokens: 262144, knowledgeCutoff: '2025-04', structuredOutput: true, lastUpdated: '2025-04', @@ -13280,7 +14158,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, - maxOutputTokens: 16384, + maxOutputTokens: 32768, knowledgeCutoff: '2025-03-31', structuredOutput: true, lastUpdated: '2025-10-06', @@ -13342,7 +14220,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 262144, - maxOutputTokens: 81920, + maxOutputTokens: 262144, structuredOutput: true, lastUpdated: '2026-02-23', capabilities: { vision: true, reasoning: true, functionCalling: true }, @@ -13549,6 +14427,33 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { toggle: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'qwen/qwen3.8-2.4t-a95b': { + displayName: 'Qwen3.8 2.4T A95B', + description: + 'Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 1048576, + maxOutputTokens: 262144, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'qwen/qwen3.8-27b': { + displayName: 'Qwen3.8 27B', + description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 1000000, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'qwen/qwen3.8-max': { displayName: 'Qwen3.8 Max', description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', @@ -13625,6 +14530,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high', 'xhigh', 'max'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'sakana/sakana-namazu': { + displayName: 'Sakana Namazu', + description: 'Multi-agent model for routing expert agents across complex analytical tasks', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 262144, + maxOutputTokens: 65536, + structuredOutput: true, + lastUpdated: '2026-08-03', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'sao10k/l3-lunaris-8b': { displayName: 'Llama 3 8B Lunaris', description: 'Open Llama instruction model for multilingual chat, reasoning, and coding', @@ -13793,7 +14711,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 1048576, - maxOutputTokens: 1048576, + maxOutputTokens: 262144, structuredOutput: false, lastUpdated: '2026-07-15', capabilities: { vision: true, reasoning: true, functionCalling: true }, @@ -13807,7 +14725,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://openrouter.ai/models', contextWindow: 524288, maxOutputTokens: 262144, - structuredOutput: false, + structuredOutput: true, lastUpdated: '2026-07-30', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'max'] }, @@ -13820,7 +14738,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 6144, - maxOutputTokens: 2048, + maxOutputTokens: 6144, knowledgeCutoff: '2023-06-30', structuredOutput: true, lastUpdated: '2023-07-22', @@ -13839,6 +14757,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'upstage/solar-pro4': { + displayName: 'Solar Pro 4', + description: 'Flagship model for demanding analysis, coding, and production agent workflows', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 524288, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-10', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'writer/palmyra-x5': { displayName: 'Palmyra X5', description: 'General-purpose chat model for instruction following, writing, and analysis', @@ -13894,8 +14824,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'x-ai/grok-4.5': { displayName: 'Grok 4.5', - description: - "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk", + description: "xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk", lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 500000, @@ -13906,6 +14835,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'x-ai/grok-4.6': { + displayName: 'Grok 4.6', + description: 'Grok model for agentic tool use, reasoning, coding, and live assistance', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 500000, + maxOutputTokens: 500000, + knowledgeCutoff: '2026-02-01', + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'x-ai/grok-build-0.1': { displayName: 'Grok Build 0.1', description: 'Fast Grok coding model tuned for agentic engineering and iterative edits', @@ -14047,7 +14990,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 204800, - maxOutputTokens: 131072, + maxOutputTokens: 128000, structuredOutput: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, @@ -14072,7 +15015,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://openrouter.ai/models', contextWindow: 204800, - maxOutputTokens: 131072, + maxOutputTokens: 128000, structuredOutput: true, lastUpdated: '2026-04-07', capabilities: { vision: false, reasoning: true, functionCalling: true }, @@ -14089,7 +15032,34 @@ export const GENERATED_MODELS_DEV_METADATA: Record< structuredOutput: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, - thinkingOptions: { efforts: ['high', 'xhigh'] }, + thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'z-ai/glm-5.2:free': { + displayName: 'GLM 5.2 (free)', + description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 256000, + maxOutputTokens: 256000, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-06-13', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'z-ai/glm-5.3': { + displayName: 'GLM-5.3', + description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering', + lifecycle: 'active', + docsUrl: 'https://openrouter.ai/models', + contextWindow: 1048576, + maxOutputTokens: 131072, + structuredOutput: false, + lastUpdated: '2026-08-14', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'] }, modalities: { input: ['text'], output: ['text'] }, }, 'z-ai/glm-5v-turbo': { @@ -15096,6 +16066,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high', 'max'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek-ai/DeepSeek-V4-Pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: + 'DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes', + lifecycle: 'active', + docsUrl: 'https://docs.together.ai/docs/serverless-models', + contextWindow: 1048576, + maxOutputTokens: 384000, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'essentialai/Rnj-1-Instruct': { displayName: 'Rnj-1 Instruct', description: @@ -15453,7 +16437,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'Open flagship GLM for long-horizon coding agents and million-token context work', lifecycle: 'active', docsUrl: 'https://docs.together.ai/docs/serverless-models', - contextWindow: 262144, + contextWindow: 512000, maxOutputTokens: 164000, structuredOutput: true, lastUpdated: '2026-06-16', @@ -15470,6 +16454,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 202752, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-02-11', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -15482,6 +16467,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-03-08', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15493,6 +16479,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-03-08', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15504,6 +16491,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-03-08', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15515,6 +16503,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-03-08', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15527,6 +16516,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 262144, maxOutputTokens: 32768, knowledgeCutoff: '2025-01', + isFree: true, lastUpdated: '2026-01-27', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -15539,6 +16529,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 204800, maxOutputTokens: 32768, + isFree: true, lastUpdated: '2026-02-12', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15550,6 +16541,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1772/128947', contextWindow: 131072, maxOutputTokens: 16384, + isFree: true, lastUpdated: '2026-03-08', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15563,6 +16555,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1823/130060', contextWindow: 256000, maxOutputTokens: 64000, + isFree: true, lastUpdated: '2026-07-06', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, @@ -15577,6 +16570,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1823/130050', contextWindow: 256000, maxOutputTokens: 64000, + isFree: true, lastUpdated: '2026-07-06', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, @@ -15589,6 +16583,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://cloud.tencent.com/document/product/1823/130050', contextWindow: 256000, maxOutputTokens: 64000, + isFree: true, lastUpdated: '2026-04-20', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true }, @@ -15829,7 +16824,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 131072, maxOutputTokens: 32768, - knowledgeCutoff: '2025-09', + knowledgeCutoff: '2025-04', lastUpdated: '2025-09', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -15842,7 +16837,9 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 131072, maxOutputTokens: 129024, - lastUpdated: '2026-05-01', + knowledgeCutoff: '2025-03-31', + structuredOutput: true, + lastUpdated: '2025-09-23', capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: ['text'] }, }, @@ -15969,6 +16966,35 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { toggle: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'alibaba/qwen3.8-2.4t-a95b': { + displayName: 'Qwen3.8 2.4T A95B', + description: + 'Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 262144, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] }, + modalities: { input: ['text'], output: ['text'] }, + }, + 'alibaba/qwen3.8-27b': { + displayName: 'Qwen3.8 27B', + description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1000000, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-14', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'alibaba/qwen3.8-max': { displayName: 'Qwen 3.8 Max', description: @@ -16285,6 +17311,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'anthropic/claude-opus-5-fast': { + displayName: 'Claude Opus 5 (Fast)', + description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1000000, + maxOutputTokens: 128000, + knowledgeCutoff: '2026-05', + lastUpdated: '2026-07-24', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'anthropic/claude-sonnet-4': { displayName: 'Claude Sonnet 4', description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control', @@ -16343,7 +17383,8 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'arcee-ai/trinity-large-thinking': { displayName: 'Trinity Large Thinking', - description: 'Flagship model for demanding analysis, coding, and production agent workflows', + description: + 'Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use', lifecycle: 'active', docsUrl: 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', @@ -16355,7 +17396,8 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'arcee-ai/trinity-mini': { displayName: 'Trinity Mini', - description: 'Efficient model for low-latency assistance, extraction, and routine automation', + description: + 'Reasoning-tuned 26B MoE model with 3B active parameters for agents, tools, and multi-step workloads', lifecycle: 'active', docsUrl: 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', @@ -16550,6 +17592,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image'], output: [] }, }, + 'bytedance/seedance-2.5': { + displayName: 'Seedance 2.5', + description: 'Video model for prompt-guided generation, editing, and motion workflows', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 0, + maxOutputTokens: 0, + lastUpdated: '2026-08-07', + capabilities: { vision: false, reasoning: false, functionCalling: false }, + modalities: { input: ['text'], output: [] }, + }, 'bytedance/seedance-v1.0-pro': { displayName: 'Seedance v1.0 Pro', description: 'Video model for prompt-guided generation, editing, and motion workflows', @@ -16720,7 +17774,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 163840, maxOutputTokens: 163840, - knowledgeCutoff: '2024-07', lastUpdated: '2024-12-26', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -16733,7 +17786,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 163840, maxOutputTokens: 128000, - knowledgeCutoff: '2024-07', lastUpdated: '2025-08-21', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -16825,6 +17877,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'deepseek/deepseek-v4-pro-0813': { + displayName: 'DeepSeek V4 Pro 0813', + description: 'Flagship DeepSeek model for coding, reasoning, and agentic work', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1000000, + maxOutputTokens: 384000, + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'fish-audio/s1': { displayName: 'S1', description: 'Speech generation model for controllable voice, narration, and audio delivery', @@ -16837,6 +17903,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, }, + 'fish-audio/s1-free': { + displayName: 'S1 (Free)', + description: 'Speech generation model for controllable voice, narration, and audio delivery', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 0, + maxOutputTokens: 0, + lastUpdated: '2025-10-20', + capabilities: { vision: false, reasoning: false, functionCalling: false }, + modalities: { input: ['text'], output: ['audio'] }, + }, 'fish-audio/s2-pro': { displayName: 'S2 Pro', description: 'Speech generation model for controllable voice, narration, and audio delivery', @@ -16849,6 +17927,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, }, + 'fish-audio/s2-pro-free': { + displayName: 'S2 Pro (Free)', + description: 'Speech generation model for controllable voice, narration, and audio delivery', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 0, + maxOutputTokens: 0, + lastUpdated: '2026-03-09', + capabilities: { vision: false, reasoning: false, functionCalling: false }, + modalities: { input: ['text'], output: ['audio'] }, + }, 'fish-audio/s2.1-pro': { displayName: 'S2.1 Pro', description: 'Speech generation model for controllable voice, narration, and audio delivery', @@ -16861,6 +17951,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, }, + 'fish-audio/s2.1-pro-free': { + displayName: 'S2.1 Pro (Free)', + description: 'Speech generation model for controllable voice, narration, and audio delivery', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 0, + maxOutputTokens: 0, + lastUpdated: '2026-07-28', + capabilities: { vision: false, reasoning: false, functionCalling: false }, + modalities: { input: ['text'], output: ['audio'] }, + }, 'fish-audio/transcribe-1': { displayName: 'Transcribe-1', description: 'Speech transcription model for accurate audio-to-text and captioning workflows', @@ -16873,6 +17975,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['audio'], output: ['text'] }, }, + 'fish-audio/transcribe-1-free': { + displayName: 'Transcribe-1 (Free)', + description: 'Speech transcription model for accurate audio-to-text and captioning workflows', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 0, + maxOutputTokens: 0, + lastUpdated: '2026-03-01', + capabilities: { vision: false, reasoning: false, functionCalling: false }, + modalities: { input: ['audio'], output: ['text'] }, + }, 'google/gemini-2.5-flash': { displayName: 'Gemini 2.5 Flash', description: 'Fast Gemini workhorse for multimodal apps where latency and price matter', @@ -17075,6 +18189,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'google/gemini-3.7-flash': { + displayName: 'Gemini 3.7 Flash', + description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1000000, + maxOutputTokens: 65536, + knowledgeCutoff: '2026-03', + structuredOutput: true, + lastUpdated: '2026-08-13', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'google/gemini-embedding-001': { displayName: 'Gemini Embedding 001', description: @@ -17134,7 +18263,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lifecycle: 'active', docsUrl: 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', - contextWindow: 256000, + contextWindow: 262144, maxOutputTokens: 131072, structuredOutput: true, lastUpdated: '2026-04-02', @@ -17302,18 +18431,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, }, - 'inclusionai/ling-3.0-tiny-free': { - displayName: 'Ling 3.0 Tiny (Free)', - description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads', - lifecycle: 'active', - docsUrl: - 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', - contextWindow: 256000, - maxOutputTokens: 32000, - lastUpdated: '2026-08-06', - capabilities: { vision: false, reasoning: true, functionCalling: true }, - modalities: { input: ['text'], output: ['text'] }, - }, 'interfaze/interfaze-beta': { displayName: 'Interfaze Beta', description: 'Multimodal reasoning model for visual analysis, planning, and tool use', @@ -17511,6 +18628,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, knowledgeCutoff: '2023-12', + isFree: true, lastUpdated: '2024-12-06', capabilities: { vision: false, reasoning: false, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -17524,6 +18642,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, knowledgeCutoff: '2024-08', + isFree: true, lastUpdated: '2025-04-05', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -17537,10 +18656,27 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 128000, maxOutputTokens: 4096, knowledgeCutoff: '2024-08', + isFree: true, lastUpdated: '2025-04-05', capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'meta/muse-glimmer-30b': { + displayName: 'Muse Glimmer 30B', + description: + 'Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 131072, + maxOutputTokens: 131072, + knowledgeCutoff: '2026-01-04', + structuredOutput: true, + lastUpdated: '2026-08-10', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'meta/muse-spark-1.1': { displayName: 'Muse Spark 1.1', description: 'Open Llama instruction model for multilingual chat, reasoning, and coding', @@ -17613,7 +18749,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 204800, maxOutputTokens: 131072, - knowledgeCutoff: '2024-10', lastUpdated: '2025-12-23', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -18052,7 +19187,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 262144, maxOutputTokens: 262144, - knowledgeCutoff: '2024-10', lastUpdated: '2025-12-15', capabilities: { vision: false, reasoning: true, functionCalling: false }, thinkingOptions: { toggle: true }, @@ -18085,6 +19219,20 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'nvidia/nemotron-3.5-lightning': { + displayName: 'Nemotron 3.5 Lightning 30B', + description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 262144, + maxOutputTokens: 131072, + structuredOutput: true, + lastUpdated: '2026-08-11', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { toggle: true }, + modalities: { input: ['text'], output: ['text'] }, + }, 'nvidia/nemotron-nano-12b-v2-vl': { displayName: 'Nvidia Nemotron Nano 12B V2 VL', description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows', @@ -18093,7 +19241,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 131072, maxOutputTokens: 131072, - knowledgeCutoff: '2024-10', lastUpdated: '2025-10-28', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -18107,7 +19254,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 131072, maxOutputTokens: 131072, - knowledgeCutoff: '2024-10', lastUpdated: '2025-08-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -18157,6 +19303,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-4.1-fast': { + displayName: 'GPT-4.1 (Fast)', + description: + 'Long-lived GPT workhorse for coding, instruction following, and production apps', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1047576, + inputLimit: 1014808, + maxOutputTokens: 32768, + knowledgeCutoff: '2024-04', + structuredOutput: true, + lastUpdated: '2025-04-14', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-4.1-mini': { displayName: 'GPT-4.1 mini', description: 'Affordable GPT-4.1 lane for fast coding help and structured extraction', @@ -18171,6 +19333,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-4.1-mini-fast': { + displayName: 'GPT-4.1 mini (Fast)', + description: 'Affordable GPT-4.1 lane for fast coding help and structured extraction', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1047576, + inputLimit: 1014808, + maxOutputTokens: 32768, + knowledgeCutoff: '2024-04', + structuredOutput: true, + lastUpdated: '2025-04-14', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-4.1-nano': { displayName: 'GPT-4.1 nano', description: 'Tiny GPT-4.1 option for classification, routing, and very high-volume tasks', @@ -18185,6 +19362,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'openai/gpt-4.1-nano-fast': { + displayName: 'GPT-4.1 nano (Fast)', + description: 'Tiny GPT-4.1 option for classification, routing, and very high-volume tasks', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1047576, + inputLimit: 1014808, + maxOutputTokens: 32768, + knowledgeCutoff: '2024-04', + structuredOutput: true, + lastUpdated: '2025-04-14', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-4o': { displayName: 'GPT-4o', description: 'Omni-era GPT for multimodal chat, practical coding, and general assistants', @@ -18199,6 +19391,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-4o-fast': { + displayName: 'GPT-4o (Fast)', + description: 'Omni-era GPT for multimodal chat, practical coding, and general assistants', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 128000, + inputLimit: 111616, + maxOutputTokens: 16384, + knowledgeCutoff: '2023-09', + structuredOutput: true, + lastUpdated: '2024-08-06', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-4o-mini': { displayName: 'GPT-4o mini', description: 'Small omni GPT for cheap multimodal assistance and production-scale traffic', @@ -18213,6 +19420,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-4o-mini-fast': { + displayName: 'GPT-4o mini (Fast)', + description: 'Small omni GPT for cheap multimodal assistance and production-scale traffic', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 128000, + inputLimit: 111616, + maxOutputTokens: 16384, + knowledgeCutoff: '2023-09', + structuredOutput: true, + lastUpdated: '2024-07-18', + capabilities: { vision: true, reasoning: false, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-4o-mini-search-preview': { displayName: 'GPT 4o Mini Search Preview', description: 'Compact GPT model for low-latency assistance and high-volume workloads', @@ -18284,6 +19506,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5-fast': { + displayName: 'GPT-5 (Fast)', + description: 'Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 400000, + inputLimit: 272000, + maxOutputTokens: 128000, + knowledgeCutoff: '2024-09-30', + structuredOutput: true, + lastUpdated: '2025-08-07', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5-mini': { displayName: 'GPT-5 Mini', description: 'Small GPT-5 for responsive agents, coding help, and everyday automation', @@ -18300,6 +19538,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'openai/gpt-5-mini-fast': { + displayName: 'GPT-5 mini (Fast)', + description: 'Small GPT-5 for responsive agents, coding help, and everyday automation', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 400000, + inputLimit: 272000, + maxOutputTokens: 128000, + knowledgeCutoff: '2024-05-30', + structuredOutput: true, + lastUpdated: '2025-08-07', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5-nano': { displayName: 'GPT-5 Nano', description: 'Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs', @@ -18325,7 +19579,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 400000, inputLimit: 128000, maxOutputTokens: 272000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2024-09-30', structuredOutput: true, lastUpdated: '2025-10-06', capabilities: { vision: true, reasoning: true, functionCalling: true }, @@ -18341,7 +19595,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 400000, inputLimit: 272000, maxOutputTokens: 128000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2024-09-30', structuredOutput: true, lastUpdated: '2025-11-13', capabilities: { vision: true, reasoning: true, functionCalling: true }, @@ -18358,7 +19612,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 400000, inputLimit: 272000, maxOutputTokens: 128000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2024-09-30', structuredOutput: true, lastUpdated: '2025-11-13', capabilities: { vision: true, reasoning: true, functionCalling: true }, @@ -18375,27 +19629,13 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 400000, inputLimit: 272000, maxOutputTokens: 128000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2024-09-30', structuredOutput: true, lastUpdated: '2025-11-13', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, - 'openai/gpt-5.1-instant': { - displayName: 'GPT-5.1 Instant', - description: 'Compact GPT model for low-latency assistance and high-volume workloads', - lifecycle: 'active', - docsUrl: - 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', - contextWindow: 128000, - inputLimit: 111616, - maxOutputTokens: 16384, - knowledgeCutoff: '2024-10', - lastUpdated: '2025-08-07', - capabilities: { vision: true, reasoning: false, functionCalling: true }, - modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, - }, 'openai/gpt-5.1-thinking': { displayName: 'GPT 5.1 Thinking', description: 'Image model for prompt-driven generation, editing, and visual design workflows', @@ -18411,6 +19651,19 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text', 'image'] }, }, + 'openai/gpt-5.1-thinking-fast': { + displayName: 'GPT 5.1 Thinking (Fast)', + description: 'Compact GPT model for low-latency assistance and high-volume workloads', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 400000, + inputLimit: 272000, + maxOutputTokens: 128000, + lastUpdated: '2025-11-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.2': { displayName: 'GPT-5.2', description: @@ -18438,13 +19691,30 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 400000, inputLimit: 272000, maxOutputTokens: 128000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2025-08-31', structuredOutput: true, lastUpdated: '2025-12-11', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.2-fast': { + displayName: 'GPT 5.2 (Fast)', + description: + 'Reliable GPT generation for broad coding, writing, and tool-assisted product work', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 400000, + inputLimit: 272000, + maxOutputTokens: 128000, + knowledgeCutoff: '2025-08-31', + structuredOutput: true, + lastUpdated: '2025-12-11', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.2-pro': { displayName: 'GPT 5.2 ', description: 'Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows', @@ -18454,29 +19724,32 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 400000, inputLimit: 272000, maxOutputTokens: 128000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2025-08-31', structuredOutput: false, lastUpdated: '2025-12-11', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, - 'openai/gpt-5.3-chat': { - displayName: 'GPT-5.3 Chat', + 'openai/gpt-5.3-codex': { + displayName: 'GPT 5.3 Codex', description: - 'Chat-tuned GPT model for conversational assistance, writing, and tool workflows', + 'Coding-optimized GPT model for repository edits, reviews, and agentic software work', lifecycle: 'active', docsUrl: 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', - contextWindow: 128000, - inputLimit: 111616, - maxOutputTokens: 16384, - lastUpdated: '2026-03-06', - capabilities: { vision: true, reasoning: false, functionCalling: true }, + contextWindow: 400000, + inputLimit: 272000, + maxOutputTokens: 128000, + knowledgeCutoff: '2025-08-31', + structuredOutput: true, + lastUpdated: '2026-02-05', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, - 'openai/gpt-5.3-codex': { - displayName: 'GPT 5.3 Codex', + 'openai/gpt-5.3-codex-fast': { + displayName: 'GPT 5.3 Codex (Fast)', description: 'Coding-optimized GPT model for repository edits, reviews, and agentic software work', lifecycle: 'active', @@ -18508,6 +19781,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.4-fast': { + displayName: 'GPT 5.4 (Fast)', + description: 'Agent-ready GPT for coding and computer-use workflows at a lower cost', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1050000, + inputLimit: 922000, + maxOutputTokens: 128000, + knowledgeCutoff: '2025-08-31', + structuredOutput: true, + lastUpdated: '2026-03-05', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.4-mini': { displayName: 'GPT 5.4 Mini', description: 'Strong small GPT for coding subagents, quick tool use, and high-volume work', @@ -18524,6 +19813,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.4-mini-fast': { + displayName: 'GPT 5.4 Mini (Fast)', + description: 'Strong small GPT for coding subagents, quick tool use, and high-volume work', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 400000, + inputLimit: 272000, + maxOutputTokens: 128000, + knowledgeCutoff: '2025-08-31', + structuredOutput: true, + lastUpdated: '2026-03-17', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.4-nano': { displayName: 'GPT 5.4 Nano', description: 'Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation', @@ -18572,6 +19877,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.5-fast': { + displayName: 'GPT 5.5 (Fast)', + description: 'Default frontier GPT for coding, computer use, research, and knowledge work', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1000000, + inputLimit: 872000, + maxOutputTokens: 128000, + knowledgeCutoff: '2025-12-01', + structuredOutput: true, + lastUpdated: '2026-04-23', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.5-pro': { displayName: 'GPT 5.5 Pro', description: 'Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding', @@ -18604,6 +19925,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.6-luna-fast': { + displayName: 'GPT 5.6 Luna (Fast)', + description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1050000, + inputLimit: 922000, + maxOutputTokens: 128000, + knowledgeCutoff: '2026-02-16', + structuredOutput: true, + lastUpdated: '2026-07-09', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.6-sol': { displayName: 'GPT 5.6 Sol', description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks', @@ -18620,6 +19957,23 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.6-sol-fast': { + displayName: 'GPT 5.6 Sol (Fast)', + description: + 'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1050000, + inputLimit: 922000, + maxOutputTokens: 128000, + knowledgeCutoff: '2026-02-16', + structuredOutput: true, + lastUpdated: '2026-07-09', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-5.6-terra': { displayName: 'GPT 5.6 Terra', description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks', @@ -18636,6 +19990,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/gpt-5.6-terra-fast': { + displayName: 'GPT 5.6 Terra (Fast)', + description: 'Balanced GPT-5.6 model for capable, cost-efficient everyday work', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1050000, + inputLimit: 922000, + maxOutputTokens: 128000, + knowledgeCutoff: '2026-02-16', + structuredOutput: true, + lastUpdated: '2026-07-09', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/gpt-image-1': { displayName: 'GPT Image 1', description: @@ -18838,12 +20208,28 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, inputLimit: 100000, maxOutputTokens: 100000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2024-05', lastUpdated: '2024-06-26', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['medium'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'openai/o3-fast': { + displayName: 'o3 (Fast)', + description: 'Deliberate o-series reasoner for hard math, coding, and multi-step analysis', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 200000, + inputLimit: 100000, + maxOutputTokens: 100000, + knowledgeCutoff: '2024-05', + structuredOutput: true, + lastUpdated: '2025-04-16', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/o3-mini': { displayName: 'o3-mini', description: 'Smaller o-series reasoner for economical coding, math, and planning tasks', @@ -18890,6 +20276,22 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'openai/o4-mini-fast': { + displayName: 'o4-mini (Fast)', + description: 'Fast o-series model for compact reasoning, coding, and tool use', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 200000, + inputLimit: 100000, + maxOutputTokens: 100000, + knowledgeCutoff: '2024-05', + structuredOutput: true, + lastUpdated: '2025-04-16', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'openai/text-embedding-3-large': { displayName: 'text-embedding-3-large', description: @@ -19057,6 +20459,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 256000, maxOutputTokens: 32768, structuredOutput: false, + isFree: true, lastUpdated: '2026-07-21', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -19195,6 +20598,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, }, + 'sakana/namazu': { + displayName: 'Sakana Namazu', + description: 'Multi-agent model for routing expert agents across complex analytical tasks', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 256000, + maxOutputTokens: 256000, + lastUpdated: '2026-08-03', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'stepfun/step-3.5-flash': { displayName: 'StepFun 3.5 Flash', description: 'StepFun flash lane for quick multimodal reasoning and coding assistance', @@ -19541,6 +20956,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'xai/grok-4.6': { + displayName: 'Grok 4.6', + description: 'Grok model for agentic tool use, reasoning, coding, and live assistance', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 500000, + maxOutputTokens: 500000, + knowledgeCutoff: '2026-02-01', + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, + }, 'xai/grok-build-0.1': { displayName: 'Grok Build 0.1', description: 'Fast Grok coding model tuned for agentic engineering and iterative edits', @@ -19566,6 +20996,18 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['text', 'image'] }, }, + 'xai/grok-imagine-image-2.0': { + displayName: 'Grok Imagine Image 2.0', + description: 'Image model for prompt-driven generation, editing, and visual design workflows', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 0, + maxOutputTokens: 0, + lastUpdated: '2026-08-07', + capabilities: { vision: false, reasoning: false, functionCalling: false }, + modalities: { input: ['text'], output: ['image'] }, + }, 'xai/grok-imagine-video': { displayName: 'Grok Imagine', description: 'Image model for prompt-driven generation, editing, and visual design workflows', @@ -19662,7 +21104,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< lastUpdated: '2026-04-22', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, - modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + modalities: { input: ['text', 'image'], output: ['text'] }, }, 'xiaomi/mimo-v2.5-pro': { displayName: 'MiMo V2.5 Pro', @@ -19734,34 +21176,6 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, - 'zai/glm-4.6v': { - displayName: 'GLM-4.6V', - description: 'GLM vision model for visual reasoning, documents, and multimodal agents', - lifecycle: 'active', - docsUrl: - 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', - contextWindow: 128000, - maxOutputTokens: 24000, - knowledgeCutoff: '2024-10', - lastUpdated: '2025-12-08', - capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { toggle: true }, - modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, - }, - 'zai/glm-4.6v-flash': { - displayName: 'GLM-4.6V-Flash', - description: 'GLM vision model for visual reasoning, documents, and multimodal agents', - lifecycle: 'active', - docsUrl: - 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', - contextWindow: 128000, - maxOutputTokens: 24000, - knowledgeCutoff: '2024-10', - lastUpdated: '2025-09-30', - capabilities: { vision: true, reasoning: true, functionCalling: true }, - thinkingOptions: { toggle: true }, - modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, - }, 'zai/glm-4.7': { displayName: 'GLM 4.7', description: 'Mature GLM model for dependable coding, reasoning, and structured agent tasks', @@ -19770,7 +21184,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', contextWindow: 200000, maxOutputTokens: 120000, - knowledgeCutoff: '2024-10', + knowledgeCutoff: '2025-04', lastUpdated: '2025-12-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -19876,6 +21290,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true }, modalities: { input: ['text'], output: ['text'] }, }, + 'zai/glm-5.3': { + displayName: 'GLM 5.3', + description: + 'Flagship GLM model for long-horizon coding, agents, and complex project delivery', + lifecycle: 'active', + docsUrl: + 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway', + contextWindow: 1000000, + maxOutputTokens: 12800, + structuredOutput: true, + lastUpdated: '2026-08-14', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'] }, + modalities: { input: ['text'], output: ['text'] }, + }, 'zai/glm-5v-turbo': { displayName: 'GLM 5V Turbo', description: 'Fast GLM vision model for screenshots, documents, and multimodal agent tasks', @@ -19943,8 +21372,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'grok-4.5': { displayName: 'Grok 4.5', - description: - "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk", + description: "xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk", lifecycle: 'active', docsUrl: 'https://docs.x.ai/docs/models', contextWindow: 500000, @@ -19955,6 +21383,21 @@ export const GENERATED_MODELS_DEV_METADATA: Record< thinkingOptions: { efforts: ['low', 'medium', 'high'] }, modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, }, + 'grok-4.6': { + displayName: 'Grok 4.6', + description: + "xAI's frontier model for long-running agents, coding, knowledge work, and visual projects", + lifecycle: 'active', + docsUrl: 'https://docs.x.ai/docs/models', + contextWindow: 500000, + maxOutputTokens: 500000, + knowledgeCutoff: '2026-02-01', + structuredOutput: true, + lastUpdated: '2026-08-12', + capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] }, + modalities: { input: ['text', 'image', 'pdf'], output: ['text'] }, + }, 'grok-build-0.1': { displayName: 'Grok Build 0.1', description: 'Fast Grok coding model tuned for agentic engineering and iterative edits', @@ -19978,6 +21421,17 @@ export const GENERATED_MODELS_DEV_METADATA: Record< capabilities: { vision: true, reasoning: false, functionCalling: false }, modalities: { input: ['text', 'image', 'pdf'], output: ['image'] }, }, + 'grok-imagine-image-2.0': { + displayName: 'Grok Imagine Image 2.0', + description: 'Image model for prompt-driven generation, editing, and visual design workflows', + lifecycle: 'active', + docsUrl: 'https://docs.x.ai/docs/models', + contextWindow: 8000, + maxOutputTokens: 0, + lastUpdated: '2026-08-07', + capabilities: { vision: true, reasoning: false, functionCalling: false }, + modalities: { input: ['text', 'image', 'pdf'], output: ['image'] }, + }, 'grok-imagine-image-quality': { displayName: 'Grok Imagine Image Quality', description: 'Image model for prompt-driven generation, editing, and visual design workflows', @@ -20101,6 +21555,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20113,6 +21568,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20125,6 +21581,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20138,6 +21595,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20150,6 +21608,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20161,6 +21620,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20172,6 +21632,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20186,6 +21647,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20198,6 +21660,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20210,6 +21673,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20223,6 +21687,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20235,6 +21700,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20246,6 +21712,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20257,6 +21724,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20271,6 +21739,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20283,6 +21752,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-03-18', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20295,6 +21765,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20308,6 +21779,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, knowledgeCutoff: '2024-12', + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20320,6 +21792,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20331,6 +21804,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20342,6 +21816,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< docsUrl: 'https://platform.xiaomimimo.com/#/docs', contextWindow: 8192, maxOutputTokens: 8192, + isFree: true, lastUpdated: '2026-04-22', capabilities: { vision: false, reasoning: false, functionCalling: false }, modalities: { input: ['text'], output: ['audio'] }, @@ -20382,6 +21857,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 131072, maxOutputTokens: 98304, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-07-28', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20447,6 +21923,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 131072, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2026-01-19', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20541,6 +22018,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 204800, maxOutputTokens: 131072, knowledgeCutoff: '2025-04', + isFree: true, lastUpdated: '2025-12-22', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20554,6 +22032,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-03-16', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { toggle: true }, @@ -20568,6 +22047,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'] }, @@ -20582,18 +22062,34 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'] }, modalities: { input: ['text'], output: ['text'] }, }, + 'glm-5.3': { + displayName: 'GLM-5.3', + description: + 'Flagship GLM model for long-horizon coding, agents, and complex project delivery', + lifecycle: 'active', + docsUrl: 'https://docs.z.ai/devpack/overview', + contextWindow: 1000000, + maxOutputTokens: 131072, + structuredOutput: true, + isFree: true, + lastUpdated: '2026-08-14', + capabilities: { vision: false, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'high', 'max'] }, + modalities: { input: ['text'], output: ['text'] }, + }, }, zenmux: { 'anthropic/claude-3.5-haiku': { displayName: 'Claude 3.5 Haiku', description: 'Fast Claude model for responsive assistance, classification, and lightweight agents', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 200000, maxOutputTokens: 64000, @@ -20605,7 +22101,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'anthropic/claude-3.7-sonnet': { displayName: 'Claude 3.7 Sonnet', description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 200000, maxOutputTokens: 64000, @@ -20780,6 +22276,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 128000, knowledgeCutoff: '2026-01-31', + isFree: true, lastUpdated: '2026-06-30', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -20800,7 +22297,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'deepseek/deepseek-chat': { displayName: 'DeepSeek-V3.2 (Non-thinking Mode)', description: 'DeepSeek chat model for instruction following, coding, and analysis', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 128000, maxOutputTokens: 64000, @@ -20970,7 +22467,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< displayName: 'Ling-1T', description: 'Tool-capable chat model for instruction following and agentic application workflows', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 128000, maxOutputTokens: 64000, @@ -20983,7 +22480,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< displayName: 'Ring-1T', description: 'Reasoning model for deliberate analysis, multi-step problem solving, and tool use', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 128000, maxOutputTokens: 64000, @@ -21105,7 +22602,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'moonshotai/kimi-k2-0905': { displayName: 'Kimi K2 0905', description: 'Kimi model for long-context chat, coding, and agentic reasoning', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 262000, maxOutputTokens: 64000, @@ -21117,7 +22614,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'moonshotai/kimi-k2-thinking': { displayName: 'Kimi K2 Thinking', description: 'Kimi reasoning model for long-horizon research, planning, and tool use', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 262000, maxOutputTokens: 64000, @@ -21129,7 +22626,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'moonshotai/kimi-k2-thinking-turbo': { displayName: 'Kimi K2 Thinking Turbo', description: 'Kimi reasoning model for long-horizon research, planning, and tool use', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 262000, maxOutputTokens: 64000, @@ -21188,6 +22685,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< maxOutputTokens: 262144, knowledgeCutoff: '2025-01', structuredOutput: true, + isFree: true, lastUpdated: '2026-06-12', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -21215,6 +22713,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1048576, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-07-16', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['max'], toggle: true }, @@ -21621,7 +23120,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'stepfun/step-3': { displayName: 'Step-3', description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 65536, maxOutputTokens: 64000, @@ -21665,6 +23164,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< inputLimit: 256000, maxOutputTokens: 256000, knowledgeCutoff: '2026-03-01', + isFree: true, lastUpdated: '2026-05-29', capabilities: { vision: true, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['low', 'medium', 'high'] }, @@ -21751,7 +23251,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< displayName: 'Doubao-Seed-Code', description: 'Coding model for repository understanding, refactors, and agentic engineering tasks', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 256000, maxOutputTokens: 64000, @@ -21764,7 +23264,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'x-ai/grok-4': { displayName: 'Grok 4', description: 'Grok model for agentic tool use, reasoning, coding, and live assistance', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 256000, maxOutputTokens: 64000, @@ -21776,7 +23276,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'x-ai/grok-4-fast': { displayName: 'Grok 4 Fast', description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 2000000, maxOutputTokens: 64000, @@ -21789,7 +23289,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'x-ai/grok-4.1-fast': { displayName: 'Grok 4.1 Fast', description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 2000000, maxOutputTokens: 64000, @@ -21802,7 +23302,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'x-ai/grok-4.1-fast-non-reasoning': { displayName: 'Grok 4.1 Fast Non Reasoning', description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 2000000, maxOutputTokens: 64000, @@ -21851,8 +23351,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< }, 'x-ai/grok-4.5': { displayName: 'Grok 4.5', - description: - "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk", + description: "xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk", lifecycle: 'active', docsUrl: 'https://docs.zenmux.ai', contextWindow: 500000, @@ -21878,7 +23377,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< 'x-ai/grok-code-fast-1': { displayName: 'Grok Code Fast 1', description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work', - lifecycle: 'active', + lifecycle: 'deprecated', docsUrl: 'https://docs.zenmux.ai', contextWindow: 256000, maxOutputTokens: 64000, @@ -22022,6 +23521,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 64000, knowledgeCutoff: '2025-01-01', + isFree: true, lastUpdated: '2025-12-08', capabilities: { vision: true, reasoning: true, functionCalling: true }, modalities: { input: ['text', 'image'], output: ['text'] }, @@ -22046,6 +23546,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 200000, maxOutputTokens: 64000, knowledgeCutoff: '2025-01-01', + isFree: true, lastUpdated: '2026-01-19', capabilities: { vision: false, reasoning: true, functionCalling: true }, modalities: { input: ['text'], output: ['text'] }, @@ -22122,6 +23623,7 @@ export const GENERATED_MODELS_DEV_METADATA: Record< contextWindow: 1000000, maxOutputTokens: 131072, structuredOutput: true, + isFree: true, lastUpdated: '2026-06-13', capabilities: { vision: false, reasoning: true, functionCalling: true }, thinkingOptions: { efforts: ['high', 'max'] }, @@ -22240,6 +23742,7 @@ export const GENERATED_MODELS_DEV_MODEL_PROVIDER_OVERRIDES: Record< 'gemini-3.5-flash': { npm: '@ai-sdk/google' }, 'gemini-3.5-flash-lite': { npm: '@ai-sdk/google' }, 'gemini-3.6-flash': { npm: '@ai-sdk/google' }, + 'gemini-3.7-flash': { npm: '@ai-sdk/google' }, 'gpt-5': { npm: '@ai-sdk/openai' }, 'gpt-5-codex': { npm: '@ai-sdk/openai' }, 'gpt-5-nano': { npm: '@ai-sdk/openai' }, @@ -22261,9 +23764,13 @@ export const GENERATED_MODELS_DEV_MODEL_PROVIDER_OVERRIDES: Record< 'gpt-5.6-sol': { npm: '@ai-sdk/openai' }, 'gpt-5.6-terra': { npm: '@ai-sdk/openai' }, 'grok-4.5': { npm: '@ai-sdk/openai' }, + 'grok-4.6': { npm: '@ai-sdk/openai' }, + 'grok-build-0.1': { npm: '@ai-sdk/openai' }, 'minimax-m2.1-free': { npm: '@ai-sdk/anthropic' }, 'minimax-m2.5-free': { npm: '@ai-sdk/anthropic' }, 'minimax-m3-free': { npm: '@ai-sdk/anthropic' }, + 'muse-spark-1.2': { npm: '@ai-sdk/openai' }, + 'muse-spark-1.2-contributor-free': { npm: '@ai-sdk/openai' }, 'qwen3.5-plus': { npm: '@ai-sdk/anthropic' }, 'qwen3.6-plus': { npm: '@ai-sdk/anthropic' }, 'qwen3.6-plus-free': { npm: '@ai-sdk/anthropic' }, @@ -22274,11 +23781,7 @@ export const GENERATED_MODELS_DEV_MODEL_PROVIDER_OVERRIDES: Record< 'minimax-m2.5': { npm: '@ai-sdk/anthropic' }, 'minimax-m2.7': { npm: '@ai-sdk/anthropic' }, 'minimax-m3': { npm: '@ai-sdk/anthropic' }, - 'qwen3.5-plus': { npm: '@ai-sdk/anthropic' }, - 'qwen3.6-plus': { npm: '@ai-sdk/anthropic' }, - 'qwen3.7-max': { npm: '@ai-sdk/anthropic' }, - 'qwen3.7-plus': { npm: '@ai-sdk/anthropic' }, - 'qwen3.8-max': { npm: '@ai-sdk/anthropic' }, + 'muse-spark-1.2-contributor': { npm: '@ai-sdk/openai' }, }, openrouter: {}, siliconflow: {}, @@ -22702,8 +24205,10 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = }, 'amazon-bedrock': {}, ambient: { api: 'https://api.ambient.xyz/v1' }, + amd: { api: 'https://developer.amd.com.cn/radeon/api/v1' }, anthropic: {}, anyapi: { api: 'https://api.anyapi.ai/v1' }, + arcee: { api: 'https://api.arcee.ai/api/v1' }, 'atomic-chat': { api: 'http://127.0.0.1:1337/v1' }, auriko: { api: 'https://api.auriko.ai/v1' }, azure: {}, @@ -22723,9 +24228,11 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = api: 'https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/ai/v1', }, cohere: {}, + coralbricks: { api: 'https://inference.coralbricks.ai/v1' }, cortecs: { api: 'https://api.cortecs.ai/v1' }, crof: { api: 'https://crof.ai/v1' }, crossmodel: { api: 'https://api.crossmodel.ai/v1' }, + crusoe: { api: 'https://api.inference.crusoecloud.com/v1' }, daoxe: { api: 'https://daoxe.com/v1' }, databricks: { api: 'https://${DATABRICKS_HOST}/ai-gateway/mlflow/v1' }, deepinfra: {}, @@ -22734,6 +24241,8 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = dinference: { api: 'https://api.dinference.com/v1' }, drun: { api: 'https://chat.d.run/v1' }, ebcloud: { api: 'https://maas-api.ebcloud.com/v1' }, + echo: { api: 'https://echo.tracerml.ai/v1' }, + edenai: { api: 'https://api.edenai.run/v3' }, empiriolabs: { api: 'https://api.empiriolabs.ai/v1' }, evroc: { api: 'https://models.think.evroc.com/v1' }, fastrouter: { api: 'https://go.fastrouter.ai/api/v1' }, @@ -22762,10 +24271,12 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = inferx: { api: 'https://model.inferx.net/endpoints/v1' }, infomaniak: { api: 'https://api.infomaniak.com/2/ai/${INFOMANIAK_PRODUCT_ID}/openai/v1' }, 'io-net': { api: 'https://api.intelligence.io.solutions/api/v1' }, + jalapeno: { api: 'https://api.jalapeno-cloud.ai/v1' }, jiekou: { api: 'https://api.jiekou.ai/openai' }, kenari: { api: 'https://kenari.id/v1' }, kilo: { api: 'https://api.kilo.ai/api/gateway' }, 'kimi-for-coding': { api: 'https://api.kimi.com/coding/v1' }, + kosmik: { api: 'https://api.koscompute.com/v1' }, 'kuae-cloud-coding-plan': { api: 'https://coding-plan-endpoint.kuaecloud.net/v1' }, lilac: { api: 'https://api.getlilac.com/v1' }, llama: { api: 'https://api.llama.com/compat/v1/' }, @@ -22776,7 +24287,7 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = lucidquery: { api: 'https://api.lucidquery.com/v1' }, lynkr: { api: 'http://127.0.0.1:8081/v1' }, meganova: { api: 'https://api.meganova.ai/v1' }, - 'merge-gateway': {}, + 'merge-gateway': { api: 'https://api-gateway.merge.dev/v1/ai-sdk' }, meta: { api: 'https://api.meta.ai/v1' }, minimax: { api: 'https://api.minimax.io/anthropic/v1' }, 'minimax-cn': { api: 'https://api.minimaxi.com/anthropic/v1' }, @@ -22820,11 +24331,14 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = 'regolo-ai': { api: 'https://api.regolo.ai/v1' }, requesty: { api: 'https://router.requesty.ai/v1' }, 'routing-run': { api: 'https://api.routing.run/v1' }, + runinfra: { api: 'https://api.runinfra.ai/v1' }, sakana: { api: 'https://api.sakana.ai/v1' }, + 'salad-cloud': {}, 'sap-ai-core': {}, sarvam: { api: 'https://api.sarvam.ai/v1' }, scaleway: { api: 'https://api.scaleway.ai/v1' }, - scx: { api: 'https://api.scx.ai/v1' }, + 'scnet-token-plan': { api: 'https://api.scnet.cn/api/llm/v1' }, + 'scx-ai': { api: 'https://api.scx.ai/v1' }, siliconflow: { api: 'https://api.siliconflow.com/v1' }, 'siliconflow-cn': { api: 'https://api.siliconflow.cn/v1' }, 'snowflake-cortex': { @@ -22860,6 +24374,7 @@ export const GENERATED_MODELS_DEV_DIRECTORY: Record = vultr: { api: 'https://api.vultrinference.com/v1' }, 'wafer.ai': { api: 'https://pass.wafer.ai/v1' }, wandb: { api: 'https://api.inference.wandb.ai/v1' }, + watsonx: {}, xai: {}, xiaomi: { api: 'https://api.xiaomimimo.com/v1' }, 'xiaomi-token-plan-ams': { api: 'https://token-plan-ams.xiaomimimo.com/v1' }, diff --git a/packages/core/src/model-metadata.ts b/packages/core/src/model-metadata.ts index a448eaccb1..660a493bd7 100644 --- a/packages/core/src/model-metadata.ts +++ b/packages/core/src/model-metadata.ts @@ -15,6 +15,8 @@ export interface ModelMetadata { maxOutputTokens?: number; knowledgeCutoff?: string; structuredOutput?: boolean; + /** Public catalog pricing is exactly zero for both input and output. */ + isFree?: boolean; lastUpdated?: string; capabilities?: ModelInfo['capabilities']; modalities?: ModelInfo['modalities']; @@ -91,7 +93,7 @@ export function lookupModelProviderOverride( /** * The request wire a model served over the OpenAI adapter must use. * - * OpenAI's `gpt-5*` families and xAI's `grok-4.5` are served only over the + * OpenAI's `gpt-5*` families and xAI's Responses-generation Grok models are served over the * Responses API; every other model on the native OpenAI adapter uses Chat * Completions. This is the single declared source of that protocol split, * expressed through the {@link ModelInfo.apiProtocol} seam. It is consumed by @@ -104,11 +106,20 @@ export function openAiAdapterApiProtocol( const id = modelId.trim(); return (providerType === 'deepseek' && deepSeekModelSupportsResponses(id)) || /^gpt-5/i.test(id) || - ((providerType === 'xai' || providerType === 'xai-oauth') && id === 'grok-4.5') + ((providerType === 'xai' || providerType === 'xai-oauth') && xAiModelSupportsResponses(id)) ? 'openai-responses' : 'openai-chat'; } +/** xAI documents the Responses reasoning contract for Grok 4.5 and later text models. */ +function xAiModelSupportsResponses(modelId: string): boolean { + const match = /^grok-(\d+)(?:\.(\d+))?(?:$|-)/i.exec(modelId.trim()); + if (!match) return false; + const major = Number(match[1]); + const minor = Number(match[2] ?? 0); + return major > 4 || (major === 4 && minor >= 5); +} + /** DeepSeek models whose first-party API contract includes the Responses wire. */ export function deepSeekModelSupportsResponses(modelId: string): boolean { const id = modelId.trim().toLowerCase(); diff --git a/packages/core/src/model-thinking.ts b/packages/core/src/model-thinking.ts index 36f7938ae3..fb61ef4430 100644 --- a/packages/core/src/model-thinking.ts +++ b/packages/core/src/model-thinking.ts @@ -216,6 +216,14 @@ export function pruneRelayModelProfiles( */ export interface ConnectionThinkingContext { readonly providerType: ProviderType; + readonly models?: readonly { + readonly id: string; + readonly capabilities?: { readonly reasoning?: boolean }; + readonly thinkingOptions?: { + readonly efforts?: readonly string[]; + readonly toggle?: boolean; + }; + }[]; readonly relayModelProfiles?: RelayModelProfiles; } @@ -249,7 +257,7 @@ export function thinkingVariantsForConnection( ): readonly ThinkingLevel[] { const declared = relayModelProfile(connection, modelId)?.thinkingLevels; if (declared) return declared; - return thinkingVariantsForModel(connection.providerType, modelId); + return deriveThinkingChoices(thinkingOptionsForConnection(connection, modelId)); } /** @@ -281,6 +289,23 @@ export function thinkingOptionsForModel( return lookupModelMetadata(providerType, modelId).thinkingOptions; } +/** + * Resolve one model's reasoning controls from the provider inventory first, + * then the bundled catalog. An explicit live `reasoning: false` defeats stale + * static metadata; live effort values override the snapshot while + * adapter-owned off-wire knowledge remains local. + */ +export function thinkingOptionsForConnection( + connection: ConnectionThinkingContext, + modelId: string, +): ThinkingOptions | undefined { + const inventory = connection.models?.find((model) => model.id === modelId); + if (inventory?.capabilities?.reasoning === false) return undefined; + const metadata = thinkingOptionsForModel(connection.providerType, modelId); + if (!inventory?.thinkingOptions) return metadata; + return { ...metadata, ...inventory.thinkingOptions }; +} + /** * Levels a model supports, in display order. Returns an empty list for * non-reasoning models and for provider/model combinations whose reasoning diff --git a/packages/core/src/provider-registry.ts b/packages/core/src/provider-registry.ts index e4f5d24abf..250dfc8f0c 100644 --- a/packages/core/src/provider-registry.ts +++ b/packages/core/src/provider-registry.ts @@ -592,27 +592,20 @@ const opencodeGoModelIds = toolCallingModelIds( GENERATED_MODELS_DEV_METADATA['opencode-go'], ['minimax-m3'], ).filter((id) => GENERATED_MODELS_DEV_METADATA['opencode-go'][id]?.lifecycle !== 'deprecated'); -// opencode-free is Maka's first-class free anonymous default. It shares the -// OpenCode Zen endpoint and model ids, but exposes only the active free -// (cost.input === 0) models from the models.dev opencode snapshot. The -// snapshot carries no cost field, so the free set is pinned here; each id is -// validated against the opencode snapshot for active + tool-capable, mirroring -// the bootstrap validation every other opencode plan entry performs. -const opencodeFreeModelIds = [ - OPENCODE_FREE_DEFAULT_MODEL, - 'mimo-v2.5-free', - 'big-pickle', - 'deepseek-v4-flash-free', - 'north-mini-code-free', - 'laguna-s-2.1-free', -] as const; -for (const id of opencodeFreeModelIds) { - const model = GENERATED_MODELS_DEV_METADATA.opencode[id]; - if (!model?.capabilities?.functionCalling || model.lifecycle === 'deprecated') { - throw new Error( - `models.dev opencode snapshot is missing an active tool-capable free model ${id} for opencode-free`, - ); - } +// opencode-free shares the OpenCode Zen endpoint and derives its inventory +// from the same catalog facts as every other provider. A model enters only +// while upstream declares it active, tool-capable, and zero-cost; retirement +// therefore cannot leave a hand-maintained id that breaks application startup. +const opencodeFreeModelIds = Object.entries(GENERATED_MODELS_DEV_METADATA.opencode) + .filter( + ([, model]) => + model.isFree && model.capabilities?.functionCalling && model.lifecycle !== 'deprecated', + ) + .map(([id]) => id); +if (!opencodeFreeModelIds.includes(OPENCODE_FREE_DEFAULT_MODEL)) { + throw new Error( + `models.dev opencode snapshot is missing free default ${OPENCODE_FREE_DEFAULT_MODEL}`, + ); } const githubCopilot = GENERATED_MODELS_DEV_PROVIDER_FACTS['github-copilot']; if (githubCopilot.id !== 'github-copilot') { diff --git a/packages/core/src/runtime-policy/connection-catalog-codec.ts b/packages/core/src/runtime-policy/connection-catalog-codec.ts index 9164336979..78994f0f57 100644 --- a/packages/core/src/runtime-policy/connection-catalog-codec.ts +++ b/packages/core/src/runtime-policy/connection-catalog-codec.ts @@ -415,7 +415,15 @@ export function decodeConnectionModel(value: unknown): ConnectionModel { const item = exactRecord( value, 'connection model', - ['id', 'displayName', 'apiProtocol', 'contextWindow', 'maxOutputTokens', 'capabilities'], + [ + 'id', + 'displayName', + 'apiProtocol', + 'contextWindow', + 'maxOutputTokens', + 'capabilities', + 'thinkingOptions', + ], ['id'], ); if ( @@ -442,6 +450,31 @@ export function decodeConnectionModel(value: unknown): ConnectionModel { ); } } + let thinkingOptions: ConnectionModel['thinkingOptions']; + if (item.thinkingOptions !== undefined) { + const raw = exactRecord( + item.thinkingOptions, + 'connection model thinking options', + ['efforts', 'toggle'], + [], + ); + thinkingOptions = {}; + if (raw.efforts !== undefined) { + if ( + !Array.isArray(raw.efforts) || + raw.efforts.length > 32 || + raw.efforts.some( + (effort) => typeof effort !== 'string' || effort.length === 0 || effort.length > 64, + ) + ) { + throw domainError('connection model thinking efforts are invalid'); + } + thinkingOptions.efforts = raw.efforts as string[]; + } + if (raw.toggle !== undefined) { + thinkingOptions.toggle = booleanValue(raw.toggle, 'connection model thinking toggle'); + } + } return { id: decodeConnectionModelId(item.id), ...(item.displayName === undefined @@ -469,6 +502,7 @@ export function decodeConnectionModel(value: unknown): ConnectionModel { ), }), ...(capabilities === undefined ? {} : { capabilities }), + ...(thinkingOptions === undefined ? {} : { thinkingOptions }), }; } diff --git a/packages/runtime/src/__tests__/model-factory-thinking.test.ts b/packages/runtime/src/__tests__/model-factory-thinking.test.ts index 919ac5e78c..7a35300d8a 100644 --- a/packages/runtime/src/__tests__/model-factory-thinking.test.ts +++ b/packages/runtime/src/__tests__/model-factory-thinking.test.ts @@ -133,15 +133,15 @@ describe('buildProviderOptions: thinking level', () => { }); }); - test('google effort model (gemini-3) sends thinkingLevel; Gemini 2.5 Flash off sends thinkingBudget 0; safetySettings always present', () => { - const g3 = buildProviderOptions(conn('google'), 'gemini-3-pro-preview', 'high'); + test('Google effort models send thinkingLevel; Gemini 2.5 Flash off sends thinkingBudget 0; safetySettings always present', () => { + const g3 = buildProviderOptions(conn('google'), 'gemini-3.1-pro-preview', 'high'); assert.equal( (g3.google as { thinkingConfig: { thinkingLevel: string } }).thinkingConfig.thinkingLevel, 'high', ); assert.ok((g3.google as { safetySettings: unknown[] }).safetySettings.length > 0); - // off not in gemini-3-pro-preview variants (only low/high) → dropped → no thinkingConfig - const g3off = buildProviderOptions(conn('google'), 'gemini-3-pro-preview', 'off'); + // off is not in the model's effort variants → dropped → no thinkingConfig + const g3off = buildProviderOptions(conn('google'), 'gemini-3.1-pro-preview', 'off'); assert.equal((g3off.google as { thinkingConfig?: unknown }).thinkingConfig, undefined); // gemini-2.5-flash is toggle-only (off); off is the Google budget-zero wire. const g25 = buildProviderOptions(conn('google'), 'gemini-2.5-flash', 'off'); @@ -268,6 +268,20 @@ describe('buildProviderOptions: thinking level', () => { }); }); + test('xAI Responses-generation models share one reasoning wire contract', () => { + for (const modelId of ['grok-4.5', 'grok-4.6']) { + assert.deepEqual(buildProviderOptions(conn('xai'), modelId, 'high'), { + openai: { + store: false, + reasoningSummary: null, + include: ['reasoning.encrypted_content'], + forceReasoning: true, + reasoningEffort: 'high', + }, + }); + } + }); + test('github-copilot routes thinking by the account-declared model protocol', () => { const anthropic = { ...conn('github-copilot'), diff --git a/packages/runtime/src/__tests__/provider-contract-overrides.ts b/packages/runtime/src/__tests__/provider-contract-overrides.ts index b1371bc96d..a5c4ec530d 100644 --- a/packages/runtime/src/__tests__/provider-contract-overrides.ts +++ b/packages/runtime/src/__tests__/provider-contract-overrides.ts @@ -205,6 +205,7 @@ async function runGitHubCopilotDiscovery(): Promise { maxOutputTokens: 128_000, apiProtocol: 'openai-responses', capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, }, { id: 'claude-sonnet-4.6', @@ -213,6 +214,7 @@ async function runGitHubCopilotDiscovery(): Promise { maxOutputTokens: 128_000, apiProtocol: 'anthropic-messages', capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, }, { id: 'gemini-3.1-pro-preview', @@ -221,6 +223,7 @@ async function runGitHubCopilotDiscovery(): Promise { maxOutputTokens: 128_000, apiProtocol: 'openai-chat', capabilities: { vision: true, reasoning: true, functionCalling: true }, + thinkingOptions: { efforts: ['low', 'medium', 'high'] }, }, ]); } diff --git a/packages/runtime/src/model-factory.ts b/packages/runtime/src/model-factory.ts index f6dc871cfb..aa45ca8795 100644 --- a/packages/runtime/src/model-factory.ts +++ b/packages/runtime/src/model-factory.ts @@ -17,7 +17,7 @@ import type { ProviderRuntimeAdapter } from '@maka/core/llm-connections'; import type { ThinkingLevel } from '@maka/core/model-thinking'; import { resolveThinkingLevel, - thinkingOptionsForModel, + thinkingOptionsForConnection, thinkingVariantsForConnection, type ThinkingOptions, } from '@maka/core/model-thinking'; @@ -361,7 +361,7 @@ export function buildProviderOptions( modelId: string, thinkingLevel?: ThinkingLevel, ): SharedV4ProviderOptions { - const thinkingOptions = thinkingOptionsForModel(connection.providerType, modelId); + const thinkingOptions = thinkingOptionsForConnection(connection, modelId); const level = resolveThinkingLevel(connection, modelId, thinkingLevel); switch (connection.providerType) { case 'kimi-coding-plan': { @@ -450,22 +450,6 @@ export function buildProviderOptions( forceReasoning: true, }, }; - case 'xai': - case 'xai-oauth': - // Only grok-4.5 needs the Responses reasoning extras; every other xAI - // model serves the plain OpenAI-compatible chat wire handled below. - if (modelId === 'grok-4.5') { - return { - openai: { - store: false, - forceReasoning: true, - reasoningSummary: null, - include: ['reasoning.encrypted_content'], - ...(level ? { reasoningEffort: level } : {}), - }, - }; - } - return buildFamilyWire(connection, modelId, level, thinkingOptions); case 'volcengine-ark': return { [connection.providerType]: { @@ -546,6 +530,12 @@ function buildFamilyWire( return { openai: { store: false, + ...(connection.providerType === 'xai' || connection.providerType === 'xai-oauth' + ? { + reasoningSummary: null, + include: ['reasoning.encrypted_content'], + } + : {}), ...(reasons || reasoningReplay.contract.reasoningReplay === 'encrypted-content' ? { forceReasoning: true } : {}), diff --git a/packages/runtime/src/model-fetcher.ts b/packages/runtime/src/model-fetcher.ts index 7a599438dc..39f3d7adf8 100644 --- a/packages/runtime/src/model-fetcher.ts +++ b/packages/runtime/src/model-fetcher.ts @@ -542,6 +542,9 @@ function toGitHubCopilotModelInfo(model: RawGitHubCopilotModel): ModelInfo[] { : {}), apiProtocol, capabilities: { vision, reasoning, functionCalling: true }, + ...(supports.reasoning_effort?.length + ? { thinkingOptions: { efforts: [...new Set(supports.reasoning_effort)] } } + : {}), }, ]; } diff --git a/packages/runtime/src/telemetry/model-pricing.generated.ts b/packages/runtime/src/telemetry/model-pricing.generated.ts index 5d8e24da0e..1f6606a66b 100644 --- a/packages/runtime/src/telemetry/model-pricing.generated.ts +++ b/packages/runtime/src/telemetry/model-pricing.generated.ts @@ -94,6 +94,19 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.2, cacheWriteUsdPer1M: 2.5, }, + { + modelKey: 'alibaba:deepseek-v4-flash-0731', + inputUsdPer1M: 0.2, + outputUsdPer1M: 0.4, + cacheReadUsdPer1M: 0.04, + }, + { + modelKey: 'alibaba:glm-5.2', + inputUsdPer1M: 1.4, + outputUsdPer1M: 4.4, + cacheReadUsdPer1M: 0.28, + cacheWriteUsdPer1M: 0, + }, { modelKey: 'alibaba:qvq-max', inputUsdPer1M: 1.2, outputUsdPer1M: 4.8 }, { modelKey: 'alibaba:qwen-flash', inputUsdPer1M: 0.05, outputUsdPer1M: 0.4 }, { modelKey: 'alibaba:qwen-max', inputUsdPer1M: 1.6, outputUsdPer1M: 6.4 }, @@ -167,6 +180,8 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheWriteUsdPer1M: 2.5, }, { modelKey: 'alibaba:qwq-plus', inputUsdPer1M: 0.8, outputUsdPer1M: 2.4 }, + { modelKey: 'cerebras:gemma-4-31b', inputUsdPer1M: 0.99, outputUsdPer1M: 1.49 }, + { modelKey: 'cerebras:gpt-oss-120b', inputUsdPer1M: 0.35, outputUsdPer1M: 0.75 }, { modelKey: 'cohere:command-a-03-2025', inputUsdPer1M: 2.5, outputUsdPer1M: 10 }, { modelKey: 'cohere:command-a-plus-05-2026', inputUsdPer1M: 2.5, outputUsdPer1M: 10 }, { modelKey: 'cohere:command-a-reasoning-08-2025', inputUsdPer1M: 2.5, outputUsdPer1M: 10 }, @@ -187,6 +202,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.497, outputUsdPer1M: 4.881, }, + { + modelKey: 'cloudflare-workers-ai:@cf/deepseek-ai/deepseek-v4-flash-0731', + inputUsdPer1M: 0.44, + outputUsdPer1M: 1.32, + cacheReadUsdPer1M: 0.014, + }, + { + modelKey: 'cloudflare-workers-ai:@cf/deepseek-ai/deepseek-v4-pro-0813', + inputUsdPer1M: 1.32, + outputUsdPer1M: 3.96, + cacheReadUsdPer1M: 0.044, + }, { modelKey: 'cloudflare-workers-ai:@cf/google/gemma-4-26b-a4b-it', inputUsdPer1M: 0.1, @@ -274,6 +301,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.0509, outputUsdPer1M: 0.335, }, + { + modelKey: 'cloudflare-workers-ai:@cf/qwen/qwen3.8-27b', + inputUsdPer1M: 0.45, + outputUsdPer1M: 3.2, + cacheReadUsdPer1M: 0.05, + }, { modelKey: 'cloudflare-workers-ai:@cf/qwen/qwq-32b', inputUsdPer1M: 0.66, outputUsdPer1M: 1 }, { modelKey: 'cloudflare-workers-ai:@cf/zai-org/glm-4.7-flash', @@ -293,6 +326,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.35, }, { modelKey: 'deepinfra:deepseek-ai/DeepSeek-V3', inputUsdPer1M: 0.32, outputUsdPer1M: 0.89 }, + { + modelKey: 'deepinfra:deepseek-ai/DeepSeek-V3-0324', + inputUsdPer1M: 0.24, + outputUsdPer1M: 0.9, + cacheReadUsdPer1M: 0.135, + }, { modelKey: 'deepinfra:deepseek-ai/DeepSeek-V3.1', inputUsdPer1M: 0.25, @@ -313,9 +352,9 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'deepinfra:deepseek-ai/DeepSeek-V4-Flash-0731', - inputUsdPer1M: 0.09, + inputUsdPer1M: 0.08, outputUsdPer1M: 0.18, - cacheReadUsdPer1M: 0.018, + cacheReadUsdPer1M: 0.016, }, { modelKey: 'deepinfra:deepseek-ai/DeepSeek-V4-Pro', @@ -323,6 +362,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 2.6, cacheReadUsdPer1M: 0.1, }, + { + modelKey: 'deepinfra:deepseek-ai/DeepSeek-V4-Pro-0813', + inputUsdPer1M: 1.3, + outputUsdPer1M: 2.6, + cacheReadUsdPer1M: 0.1, + }, { modelKey: 'deepinfra:google/gemma-4-26B-A4B-it', inputUsdPer1M: 0.07, outputUsdPer1M: 0.34 }, { modelKey: 'deepinfra:google/gemma-4-31B-it', inputUsdPer1M: 0.13, outputUsdPer1M: 0.38 }, { modelKey: 'deepinfra:google/gemma-4-E4B-it', inputUsdPer1M: 0.02, outputUsdPer1M: 0.1 }, @@ -379,9 +424,9 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'deepinfra:moonshotai/Kimi-K3', - inputUsdPer1M: 2.7, - outputUsdPer1M: 13.5, - cacheReadUsdPer1M: 0.27, + inputUsdPer1M: 2.85, + outputUsdPer1M: 14.25, + cacheReadUsdPer1M: 0.285, }, { modelKey: 'deepinfra:nvidia/Llama-3.3-Nemotron-Super-49B-v1.5', @@ -406,6 +451,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.09, outputUsdPer1M: 0.55, }, + { modelKey: 'deepinfra:Qwen/Qwen3-30B-A3B', inputUsdPer1M: 0.12, outputUsdPer1M: 0.5 }, { modelKey: 'deepinfra:Qwen/Qwen3-32B', inputUsdPer1M: 0.08, outputUsdPer1M: 0.28 }, { modelKey: 'deepinfra:Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo', @@ -418,6 +464,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.09, outputUsdPer1M: 1.1, }, + { + modelKey: 'deepinfra:Qwen/Qwen3-VL-235B-A22B-Instruct', + inputUsdPer1M: 0.2, + outputUsdPer1M: 0.88, + cacheReadUsdPer1M: 0.11, + }, { modelKey: 'deepinfra:Qwen/Qwen3.5-122B-A10B', inputUsdPer1M: 0.29, outputUsdPer1M: 2.4 }, { modelKey: 'deepinfra:Qwen/Qwen3.5-27B', inputUsdPer1M: 0.26, outputUsdPer1M: 2.6 }, { @@ -435,6 +487,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'deepinfra:Qwen/Qwen3.5-9B', inputUsdPer1M: 0.1, outputUsdPer1M: 0.15 }, { modelKey: 'deepinfra:Qwen/Qwen3.6-27B', inputUsdPer1M: 0.32, outputUsdPer1M: 3.2 }, { modelKey: 'deepinfra:Qwen/Qwen3.6-35B-A3B', inputUsdPer1M: 0.1, outputUsdPer1M: 0.95 }, + { + modelKey: 'deepinfra:Qwen/Qwen3.8-2.4T-A95B', + inputUsdPer1M: 2, + outputUsdPer1M: 6, + cacheReadUsdPer1M: 0.2, + }, + { + modelKey: 'deepinfra:Qwen/Qwen3.8-27B', + inputUsdPer1M: 0.4, + outputUsdPer1M: 3, + cacheReadUsdPer1M: 0.04, + }, { modelKey: 'deepinfra:Qwen/Qwen3.8-Max', inputUsdPer1M: 1.65, @@ -555,6 +619,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 3.48, cacheReadUsdPer1M: 0.145, }, + { + modelKey: 'fireworks-ai:accounts/fireworks/models/deepseek-v4-pro-0813', + inputUsdPer1M: 1.32, + outputUsdPer1M: 3.96, + cacheReadUsdPer1M: 0.044, + }, { modelKey: 'fireworks-ai:accounts/fireworks/models/glm-5p2', inputUsdPer1M: 1.4, @@ -573,6 +643,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0.3, cacheReadUsdPer1M: 0.035, }, + { + modelKey: 'fireworks-ai:accounts/fireworks/models/inkling', + inputUsdPer1M: 1, + outputUsdPer1M: 4.05, + cacheReadUsdPer1M: 0.17, + }, { modelKey: 'fireworks-ai:accounts/fireworks/models/kimi-k2p6', inputUsdPer1M: 0.95, @@ -603,12 +679,36 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 1.2, cacheReadUsdPer1M: 0.06, }, + { + modelKey: 'fireworks-ai:accounts/fireworks/models/muse-glimmer-30b', + inputUsdPer1M: 0.35, + outputUsdPer1M: 1.5, + cacheReadUsdPer1M: 0.04, + }, + { + modelKey: 'fireworks-ai:accounts/fireworks/models/nemotron-3-ultra-nvfp4', + inputUsdPer1M: 0.6, + outputUsdPer1M: 2.4, + cacheReadUsdPer1M: 0.119, + }, + { + modelKey: 'fireworks-ai:accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b', + inputUsdPer1M: 0.05, + outputUsdPer1M: 0.2, + cacheReadUsdPer1M: 0.01, + }, { modelKey: 'fireworks-ai:accounts/fireworks/models/qwen3p7-plus', inputUsdPer1M: 0.4, outputUsdPer1M: 1.6, cacheReadUsdPer1M: 0.08, }, + { + modelKey: 'fireworks-ai:accounts/fireworks/models/qwen3p8-max', + inputUsdPer1M: 2, + outputUsdPer1M: 6, + cacheReadUsdPer1M: 0.25, + }, { modelKey: 'fireworks-ai:accounts/fireworks/routers/glm-5p2-fast', inputUsdPer1M: 2.1, @@ -639,13 +739,6 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 22.5, cacheReadUsdPer1M: 0.45, }, - { - modelKey: 'google:gemini-2.0-flash', - inputUsdPer1M: 0.1, - outputUsdPer1M: 0.4, - cacheReadUsdPer1M: 0.025, - }, - { modelKey: 'google:gemini-2.0-flash-lite', inputUsdPer1M: 0.075, outputUsdPer1M: 0.3 }, { modelKey: 'google:gemini-2.5-flash', inputUsdPer1M: 0.3, @@ -710,6 +803,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 7.5, cacheReadUsdPer1M: 0.15, }, + { + modelKey: 'google:gemini-3.7-flash', + inputUsdPer1M: 0.75, + outputUsdPer1M: 3.75, + cacheReadUsdPer1M: 0.075, + }, { modelKey: 'google:gemini-embedding-001', inputUsdPer1M: 0.15, outputUsdPer1M: 0 }, { modelKey: 'google:gemini-embedding-2', inputUsdPer1M: 0.2, outputUsdPer1M: 0 }, { @@ -728,6 +827,14 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'google:gemini-robotics-er-1.6-preview', inputUsdPer1M: 1, outputUsdPer1M: 5 }, { modelKey: 'google:lyria-3-clip-preview', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'google:lyria-3-pro-preview', inputUsdPer1M: 0, outputUsdPer1M: 0 }, + { modelKey: 'groq:allam-2-7b', inputUsdPer1M: 0, outputUsdPer1M: 0 }, + { modelKey: 'groq:llama-3.1-8b-instant', inputUsdPer1M: 0.05, outputUsdPer1M: 0.08 }, + { modelKey: 'groq:llama-3.3-70b-versatile', inputUsdPer1M: 0.59, outputUsdPer1M: 0.79 }, + { + modelKey: 'groq:meta-llama/llama-prompt-guard-2-22m', + inputUsdPer1M: 0.03, + outputUsdPer1M: 0.03, + }, { modelKey: 'groq:meta-llama/llama-prompt-guard-2-86m', inputUsdPer1M: 0.04, @@ -755,6 +862,11 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'huggingface:deepseek-ai/DeepSeek-R1', inputUsdPer1M: 0.7, outputUsdPer1M: 2.5 }, { modelKey: 'huggingface:deepseek-ai/DeepSeek-R1-0528', inputUsdPer1M: 3, outputUsdPer1M: 5 }, { modelKey: 'huggingface:deepseek-ai/DeepSeek-V3', inputUsdPer1M: 0.4, outputUsdPer1M: 1.3 }, + { + modelKey: 'huggingface:deepseek-ai/DeepSeek-V3-0324', + inputUsdPer1M: 0.27, + outputUsdPer1M: 1.12, + }, { modelKey: 'huggingface:deepseek-ai/DeepSeek-V3.1', inputUsdPer1M: 0.27, outputUsdPer1M: 1 }, { modelKey: 'huggingface:deepseek-ai/DeepSeek-V3.2', inputUsdPer1M: 0.28, outputUsdPer1M: 0.4 }, { @@ -773,8 +885,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0.87, cacheReadUsdPer1M: 0.003625, }, + { + modelKey: 'huggingface:deepseek-ai/DeepSeek-V4-Pro-0813', + inputUsdPer1M: 1.32, + outputUsdPer1M: 3.96, + }, { modelKey: 'huggingface:google/gemma-4-26B-A4B-it', inputUsdPer1M: 0.13, outputUsdPer1M: 0.4 }, { modelKey: 'huggingface:google/gemma-4-31B-it', inputUsdPer1M: 0.14, outputUsdPer1M: 0.4 }, + { + modelKey: 'huggingface:meta-llama/Llama-3.1-8B-Instruct', + inputUsdPer1M: 0.06, + outputUsdPer1M: 0.06, + }, { modelKey: 'huggingface:meta-llama/Llama-3.3-70B-Instruct', inputUsdPer1M: 0.59, @@ -819,6 +941,11 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'huggingface:moonshotai/Kimi-K3', inputUsdPer1M: 3, outputUsdPer1M: 15 }, { modelKey: 'huggingface:openai/gpt-oss-120b', inputUsdPer1M: 0.25, outputUsdPer1M: 0.69 }, { modelKey: 'huggingface:openai/gpt-oss-20b', inputUsdPer1M: 0.1, outputUsdPer1M: 0.5 }, + { + modelKey: 'huggingface:Qwen/Qwen2.5-Coder-32B-Instruct', + inputUsdPer1M: 0.06, + outputUsdPer1M: 0.2, + }, { modelKey: 'huggingface:Qwen/Qwen3-235B-A22B', inputUsdPer1M: 0.2, outputUsdPer1M: 0.8 }, { modelKey: 'huggingface:Qwen/Qwen3-235B-A22B-Instruct-2507', @@ -830,6 +957,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.3, outputUsdPer1M: 3, }, + { modelKey: 'huggingface:Qwen/Qwen3-30B-A3B', inputUsdPer1M: 0.12, outputUsdPer1M: 0.5 }, { modelKey: 'huggingface:Qwen/Qwen3-32B', inputUsdPer1M: 0.29, outputUsdPer1M: 0.59 }, { modelKey: 'huggingface:Qwen/Qwen3-Coder-30B-A3B-Instruct', @@ -854,6 +982,16 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.3, outputUsdPer1M: 2, }, + { + modelKey: 'huggingface:Qwen/Qwen3-VL-235B-A22B-Instruct', + inputUsdPer1M: 0.3, + outputUsdPer1M: 1.5, + }, + { + modelKey: 'huggingface:Qwen/Qwen3-VL-235B-A22B-Thinking', + inputUsdPer1M: 0.98, + outputUsdPer1M: 3.95, + }, { modelKey: 'huggingface:Qwen/Qwen3.5-122B-A10B', inputUsdPer1M: 0.4, outputUsdPer1M: 3.2 }, { modelKey: 'huggingface:Qwen/Qwen3.5-27B', inputUsdPer1M: 0.3, outputUsdPer1M: 2.4 }, { modelKey: 'huggingface:Qwen/Qwen3.5-35B-A3B', inputUsdPer1M: 0.25, outputUsdPer1M: 2 }, @@ -861,6 +999,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'huggingface:Qwen/Qwen3.5-9B', inputUsdPer1M: 0.17, outputUsdPer1M: 0.25 }, { modelKey: 'huggingface:Qwen/Qwen3.6-27B', inputUsdPer1M: 0.47, outputUsdPer1M: 3.19 }, { modelKey: 'huggingface:Qwen/Qwen3.6-35B-A3B', inputUsdPer1M: 0.15, outputUsdPer1M: 0.95 }, + { modelKey: 'huggingface:Qwen/Qwen3.8-2.4T-A95B', inputUsdPer1M: 2.5, outputUsdPer1M: 6.25 }, { modelKey: 'huggingface:stepfun-ai/Step-3.5-Flash', inputUsdPer1M: 0.1, outputUsdPer1M: 0.3 }, { modelKey: 'huggingface:stepfun-ai/Step-3.7-Flash', inputUsdPer1M: 0.2, outputUsdPer1M: 1.15 }, { modelKey: 'huggingface:tencent/Hy3', inputUsdPer1M: 0.14, outputUsdPer1M: 0.58 }, @@ -877,6 +1016,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'huggingface:zai-org/GLM-4.5-Air', inputUsdPer1M: 0.13, outputUsdPer1M: 0.85 }, { modelKey: 'huggingface:zai-org/GLM-4.5V', inputUsdPer1M: 0.6, outputUsdPer1M: 1.8 }, { modelKey: 'huggingface:zai-org/GLM-4.6', inputUsdPer1M: 0.55, outputUsdPer1M: 2.2 }, + { modelKey: 'huggingface:zai-org/GLM-4.6V-Flash', inputUsdPer1M: 0.3, outputUsdPer1M: 0.9 }, { modelKey: 'huggingface:zai-org/GLM-4.7', inputUsdPer1M: 0.6, @@ -1059,6 +1199,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0, }, { modelKey: 'nvidia:meta/llama-guard-4-12b', inputUsdPer1M: 0, outputUsdPer1M: 0 }, + { modelKey: 'nvidia:meta/muse-glimmer-30b', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'nvidia:microsoft/phi-4-mini-instruct', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'nvidia:microsoft/phi-4-multimodal-instruct', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'nvidia:minimaxai/minimax-m2.7', inputUsdPer1M: 0, outputUsdPer1M: 0 }, @@ -1139,6 +1280,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 2.5, cacheReadUsdPer1M: 0.15, }, + { modelKey: 'nvidia:nvidia/nemotron-3.5-lightning-30b-a3b', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'nvidia:nvidia/nemotron-content-safety-reasoning-4b', inputUsdPer1M: 0, @@ -1415,6 +1557,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 7.5, cacheReadUsdPer1M: 0.15, }, + { + modelKey: 'opencode:gemini-3.7-flash', + inputUsdPer1M: 1.5, + outputUsdPer1M: 7.5, + cacheReadUsdPer1M: 0.15, + }, { modelKey: 'opencode:glm-4.6', inputUsdPer1M: 0.6, outputUsdPer1M: 2.2, cacheReadUsdPer1M: 0.1 }, { modelKey: 'opencode:glm-4.7', inputUsdPer1M: 0.6, outputUsdPer1M: 2.2, cacheReadUsdPer1M: 0.1 }, { modelKey: 'opencode:glm-4.7-free', inputUsdPer1M: 0, outputUsdPer1M: 0, cacheReadUsdPer1M: 0 }, @@ -1587,6 +1735,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0, cacheReadUsdPer1M: 0, }, + { modelKey: 'opencode:ling-3.0-tiny-free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'opencode:longcat-2.0-free', inputUsdPer1M: 0, @@ -1659,6 +1808,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0, cacheReadUsdPer1M: 0, }, + { + modelKey: 'opencode:muse-spark-1.2', + inputUsdPer1M: 1.25, + outputUsdPer1M: 4.25, + cacheReadUsdPer1M: 0.15, + }, + { + modelKey: 'opencode:muse-spark-1.2-contributor-free', + inputUsdPer1M: 0, + outputUsdPer1M: 0, + cacheReadUsdPer1M: 0, + }, { modelKey: 'opencode:nemotron-3-super-free', inputUsdPer1M: 0, @@ -1671,6 +1832,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0, cacheReadUsdPer1M: 0, }, + { + modelKey: 'opencode:nemotron-3.5-lightning-free', + inputUsdPer1M: 0, + outputUsdPer1M: 0, + cacheReadUsdPer1M: 0, + }, { modelKey: 'opencode:north-mini-code-free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'opencode:qwen3-coder', inputUsdPer1M: 0.45, outputUsdPer1M: 1.8 }, { @@ -1725,21 +1892,21 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:~deepseek/deepseek-v4-flash-latest', - inputUsdPer1M: 0.09, - outputUsdPer1M: 0.18, - cacheReadUsdPer1M: 0.018, + inputUsdPer1M: 0.065, + outputUsdPer1M: 0.14, + cacheReadUsdPer1M: 0.014, }, { modelKey: 'openrouter:~google/gemini-flash-latest', - inputUsdPer1M: 1.5, - outputUsdPer1M: 7.5, - cacheReadUsdPer1M: 0.15, - cacheWriteUsdPer1M: 0.083333, + inputUsdPer1M: 0.375, + outputUsdPer1M: 1.875, + cacheReadUsdPer1M: 0.0375, + cacheWriteUsdPer1M: 0.020833, }, { modelKey: 'openrouter:~moonshotai/kimi-latest', - inputUsdPer1M: 2.9, - outputUsdPer1M: 14, + inputUsdPer1M: 2.6, + outputUsdPer1M: 13, cacheReadUsdPer1M: 0.29, }, { @@ -1748,7 +1915,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 4.5, cacheReadUsdPer1M: 0.075, }, - { modelKey: 'openrouter:ai21/jamba-large-1.7', inputUsdPer1M: 2, outputUsdPer1M: 8 }, + { + modelKey: 'openrouter:~z-ai/glm-latest', + inputUsdPer1M: 1.4, + outputUsdPer1M: 4.4, + cacheReadUsdPer1M: 0.26, + }, { modelKey: 'openrouter:aion-labs/aion-2.0', inputUsdPer1M: 0.8, @@ -1880,6 +2052,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.42, outputUsdPer1M: 1.25, }, + { modelKey: 'openrouter:bytedance-seed/seed-2-1-turbo', inputUsdPer1M: 0.5, outputUsdPer1M: 2.5 }, { modelKey: 'openrouter:bytedance/ui-tars-1.5-7b', inputUsdPer1M: 0.1, @@ -1902,12 +2075,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'openrouter:cohere/north-mini-code:free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'openrouter:deepcogito/cogito-v2.1-671b', inputUsdPer1M: 1.25, outputUsdPer1M: 1.25 }, { modelKey: 'openrouter:deepseek/deepseek-chat', inputUsdPer1M: 0.2574, outputUsdPer1M: 1.0287 }, - { - modelKey: 'openrouter:deepseek/deepseek-chat-v3-0324', - inputUsdPer1M: 0.27, - outputUsdPer1M: 1.12, - cacheReadUsdPer1M: 0.135, - }, + { modelKey: 'openrouter:deepseek/deepseek-chat-v3-0324', inputUsdPer1M: 0.25, outputUsdPer1M: 1 }, { modelKey: 'openrouter:deepseek/deepseek-chat-v3.1', inputUsdPer1M: 0.25, @@ -1930,7 +2098,6 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ modelKey: 'openrouter:deepseek/deepseek-v3.1-terminus', inputUsdPer1M: 0.27, outputUsdPer1M: 1, - cacheReadUsdPer1M: 0.135, }, { modelKey: 'openrouter:deepseek/deepseek-v3.2', @@ -1941,21 +2108,32 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'openrouter:deepseek/deepseek-v3.2-exp', inputUsdPer1M: 0.27, outputUsdPer1M: 0.41 }, { modelKey: 'openrouter:deepseek/deepseek-v4-flash', - inputUsdPer1M: 0.0882, - outputUsdPer1M: 0.1764, - cacheReadUsdPer1M: 0.01764, + inputUsdPer1M: 0.088606, + outputUsdPer1M: 0.177212, + cacheReadUsdPer1M: 0.017721, }, { modelKey: 'openrouter:deepseek/deepseek-v4-flash-0731', - inputUsdPer1M: 0.09, - outputUsdPer1M: 0.18, - cacheReadUsdPer1M: 0.018, + inputUsdPer1M: 0.14, + outputUsdPer1M: 0.28, + cacheReadUsdPer1M: 0.028, }, { modelKey: 'openrouter:deepseek/deepseek-v4-pro', - inputUsdPer1M: 0.435, - outputUsdPer1M: 0.87, - cacheReadUsdPer1M: 0.003625, + inputUsdPer1M: 1.6, + outputUsdPer1M: 3.2, + cacheReadUsdPer1M: 0.135, + }, + { + modelKey: 'openrouter:deepseek/deepseek-v4-pro-0813', + inputUsdPer1M: 1.188, + outputUsdPer1M: 3.564, + cacheReadUsdPer1M: 0.0396, + }, + { + modelKey: 'openrouter:dots-studio/dots-3-note-preview:free', + inputUsdPer1M: 0, + outputUsdPer1M: 0, }, { modelKey: 'openrouter:google/gemini-2.5-flash', @@ -2040,10 +2218,17 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:google/gemini-3.6-flash', - inputUsdPer1M: 1.5, - outputUsdPer1M: 7.5, - cacheReadUsdPer1M: 0.15, - cacheWriteUsdPer1M: 0.083333, + inputUsdPer1M: 0.75, + outputUsdPer1M: 3.75, + cacheReadUsdPer1M: 0.075, + cacheWriteUsdPer1M: 0.041667, + }, + { + modelKey: 'openrouter:google/gemini-3.7-flash', + inputUsdPer1M: 0.375, + outputUsdPer1M: 1.875, + cacheReadUsdPer1M: 0.0375, + cacheWriteUsdPer1M: 0.020833, }, { modelKey: 'openrouter:google/gemma-2-27b-it', inputUsdPer1M: 0.65, outputUsdPer1M: 0.65 }, { modelKey: 'openrouter:google/gemma-3-12b-it', inputUsdPer1M: 0.05, outputUsdPer1M: 0.15 }, @@ -2059,14 +2244,14 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'openrouter:google/gemma-4-26b-a4b-it:free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'openrouter:google/gemma-4-31b-it', - inputUsdPer1M: 0.1, + inputUsdPer1M: 0.09, outputUsdPer1M: 0.34, - cacheReadUsdPer1M: 0.1, + cacheReadUsdPer1M: 0.05, }, { modelKey: 'openrouter:google/gemma-4-31b-it:free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'openrouter:google/lyria-3-clip-preview', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'openrouter:google/lyria-3-pro-preview', inputUsdPer1M: 0, outputUsdPer1M: 0 }, - { modelKey: 'openrouter:gryphe/mythomax-l2-13b', inputUsdPer1M: 0.08, outputUsdPer1M: 0.11 }, + { modelKey: 'openrouter:gryphe/mythomax-l2-13b', inputUsdPer1M: 0.06, outputUsdPer1M: 0.06 }, { modelKey: 'openrouter:ibm-granite/granite-4.0-h-micro', inputUsdPer1M: 0.017, @@ -2102,7 +2287,6 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0.063, cacheReadUsdPer1M: 0.0042, }, - { modelKey: 'openrouter:inclusionai/ling-3.0-tiny:free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'openrouter:inclusionai/ring-2.6-1t', inputUsdPer1M: 0.075, @@ -2127,7 +2311,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 2.96, cacheReadUsdPer1M: 0.15, }, - { modelKey: 'openrouter:mancer/weaver', inputUsdPer1M: 0.5, outputUsdPer1M: 0.75 }, + { modelKey: 'openrouter:liquid/lfm-2.5-2.6b:free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'openrouter:meituan/longcat-2.0', inputUsdPer1M: 0.3, @@ -2167,6 +2351,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.18, outputUsdPer1M: 0.18, }, + { + modelKey: 'openrouter:meta/muse-glimmer-30b', + inputUsdPer1M: 0.35, + outputUsdPer1M: 1.5, + cacheReadUsdPer1M: 0.04, + }, { modelKey: 'openrouter:meta/muse-spark-1.1', inputUsdPer1M: 1.25, @@ -2198,15 +2388,15 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:minimax/minimax-m2.5', - inputUsdPer1M: 0.22, + inputUsdPer1M: 0.225, outputUsdPer1M: 0.9, - cacheReadUsdPer1M: 0.05, + cacheReadUsdPer1M: 0.06, }, { modelKey: 'openrouter:minimax/minimax-m2.7', - inputUsdPer1M: 0.27, - outputUsdPer1M: 1.08, - cacheReadUsdPer1M: 0.054, + inputUsdPer1M: 0.3, + outputUsdPer1M: 1.2, + cacheReadUsdPer1M: 0.06, }, { modelKey: 'openrouter:minimax/minimax-m3', @@ -2319,19 +2509,19 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:moonshotai/kimi-k2.5', - inputUsdPer1M: 0.57, - outputUsdPer1M: 2.85, - cacheReadUsdPer1M: 0.095, + inputUsdPer1M: 0.45, + outputUsdPer1M: 2.25, + cacheReadUsdPer1M: 0.07, }, { modelKey: 'openrouter:moonshotai/kimi-k2.6', - inputUsdPer1M: 0.589, - outputUsdPer1M: 2.48, - cacheReadUsdPer1M: 0.0992, + inputUsdPer1M: 0.95, + outputUsdPer1M: 4, + cacheReadUsdPer1M: 0.16, }, { modelKey: 'openrouter:moonshotai/kimi-k2.7-code', - inputUsdPer1M: 0.7, + inputUsdPer1M: 0.71, outputUsdPer1M: 3.5, cacheReadUsdPer1M: 0.15, }, @@ -2385,8 +2575,8 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:nvidia/nemotron-3-super-120b-a12b', - inputUsdPer1M: 0.3, - outputUsdPer1M: 0.9, + inputUsdPer1M: 0.085, + outputUsdPer1M: 0.4, }, { modelKey: 'openrouter:nvidia/nemotron-3-super-120b-a12b:free', @@ -2409,6 +2599,17 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0, outputUsdPer1M: 0, }, + { + modelKey: 'openrouter:nvidia/nemotron-3.5-lightning', + inputUsdPer1M: 0.08, + outputUsdPer1M: 0.2, + cacheReadUsdPer1M: 0.04, + }, + { + modelKey: 'openrouter:nvidia/nemotron-3.5-lightning:free', + inputUsdPer1M: 0, + outputUsdPer1M: 0, + }, { modelKey: 'openrouter:nvidia/nemotron-nano-12b-v2-vl:free', inputUsdPer1M: 0, @@ -2545,12 +2746,6 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.175, }, { modelKey: 'openrouter:openai/gpt-5.2-pro', inputUsdPer1M: 21, outputUsdPer1M: 168 }, - { - modelKey: 'openrouter:openai/gpt-5.3-chat', - inputUsdPer1M: 1.75, - outputUsdPer1M: 14, - cacheReadUsdPer1M: 0.175, - }, { modelKey: 'openrouter:openai/gpt-5.3-codex', inputUsdPer1M: 1.75, @@ -2583,7 +2778,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 30, cacheReadUsdPer1M: 0.5, }, - { modelKey: 'openrouter:openai/gpt-oss-120b', inputUsdPer1M: 0.037, outputUsdPer1M: 0.17 }, + { + modelKey: 'openrouter:openai/gpt-oss-120b', + inputUsdPer1M: 0.03, + outputUsdPer1M: 0.17, + cacheReadUsdPer1M: 0.03, + }, { modelKey: 'openrouter:openai/gpt-oss-20b', inputUsdPer1M: 0.03, @@ -2660,10 +2860,11 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:qwen/qwen2.5-vl-72b-instruct', - inputUsdPer1M: 0.25, - outputUsdPer1M: 0.75, + inputUsdPer1M: 0.8, + outputUsdPer1M: 1, + cacheReadUsdPer1M: 0.4, }, - { modelKey: 'openrouter:qwen/qwen3-14b', inputUsdPer1M: 0.2275, outputUsdPer1M: 0.91 }, + { modelKey: 'openrouter:qwen/qwen3-14b', inputUsdPer1M: 0.12, outputUsdPer1M: 0.24 }, { modelKey: 'openrouter:qwen/qwen3-235b-a22b', inputUsdPer1M: 0.455, outputUsdPer1M: 1.82 }, { modelKey: 'openrouter:qwen/qwen3-235b-a22b-2507', inputUsdPer1M: 0.09, outputUsdPer1M: 0.55 }, { @@ -2671,7 +2872,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ inputUsdPer1M: 0.23, outputUsdPer1M: 2.3, }, - { modelKey: 'openrouter:qwen/qwen3-30b-a3b', inputUsdPer1M: 0.12, outputUsdPer1M: 0.5 }, + { modelKey: 'openrouter:qwen/qwen3-30b-a3b', inputUsdPer1M: 0.13, outputUsdPer1M: 0.52 }, { modelKey: 'openrouter:qwen/qwen3-30b-a3b-instruct-2507', inputUsdPer1M: 0.04815, @@ -2693,7 +2894,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'openrouter:qwen/qwen3-coder-30b-a3b-instruct', inputUsdPer1M: 0.07, - outputUsdPer1M: 0.27, + outputUsdPer1M: 0.28, }, { modelKey: 'openrouter:qwen/qwen3-coder-next', @@ -2724,8 +2925,8 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:qwen/qwen3-vl-30b-a3b-instruct', - inputUsdPer1M: 0.15, - outputUsdPer1M: 0.6, + inputUsdPer1M: 0.13, + outputUsdPer1M: 0.52, }, { modelKey: 'openrouter:qwen/qwen3-vl-30b-a3b-thinking', @@ -2739,9 +2940,14 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:qwen/qwen3-vl-8b-instruct', inputUsdPer1M: 0.117, outputUsdPer1M: 0.455 }, { modelKey: 'openrouter:qwen/qwen3-vl-8b-thinking', inputUsdPer1M: 0.18, outputUsdPer1M: 2.1 }, - { modelKey: 'openrouter:qwen/qwen3.5-122b-a10b', inputUsdPer1M: 0.29, outputUsdPer1M: 2.4 }, + { modelKey: 'openrouter:qwen/qwen3.5-122b-a10b', inputUsdPer1M: 0.26, outputUsdPer1M: 2.08 }, { modelKey: 'openrouter:qwen/qwen3.5-27b', inputUsdPer1M: 0.195, outputUsdPer1M: 1.56 }, - { modelKey: 'openrouter:qwen/qwen3.5-35b-a3b', inputUsdPer1M: 0.14, outputUsdPer1M: 1 }, + { + modelKey: 'openrouter:qwen/qwen3.5-35b-a3b', + inputUsdPer1M: 0.25, + outputUsdPer1M: 1.25, + cacheReadUsdPer1M: 0.25, + }, { modelKey: 'openrouter:qwen/qwen3.5-397b-a17b', inputUsdPer1M: 0.39, outputUsdPer1M: 2.34 }, { modelKey: 'openrouter:qwen/qwen3.5-9b', inputUsdPer1M: 0.1, outputUsdPer1M: 0.15 }, { modelKey: 'openrouter:qwen/qwen3.5-flash-02-23', inputUsdPer1M: 0.065, outputUsdPer1M: 0.26 }, @@ -2764,6 +2970,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.295, cacheWriteUsdPer1M: 1.84375, }, + { + modelKey: 'openrouter:qwen/qwen3.8-2.4t-a95b', + inputUsdPer1M: 2, + outputUsdPer1M: 6, + cacheReadUsdPer1M: 0.25, + }, + { + modelKey: 'openrouter:qwen/qwen3.8-27b', + inputUsdPer1M: 0.45, + outputUsdPer1M: 3.2, + cacheReadUsdPer1M: 0.05, + }, { modelKey: 'openrouter:qwen/qwen3.8-max', inputUsdPer1M: 2, @@ -2775,6 +2993,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'openrouter:rekaai/reka-flash-3', inputUsdPer1M: 0.1, outputUsdPer1M: 0.2 }, { modelKey: 'openrouter:relace/relace-apply-3', inputUsdPer1M: 0.85, outputUsdPer1M: 1.25 }, { modelKey: 'openrouter:relace/relace-search', inputUsdPer1M: 1, outputUsdPer1M: 3 }, + { + modelKey: 'openrouter:sakana/sakana-namazu', + inputUsdPer1M: 0.95, + outputUsdPer1M: 4, + cacheReadUsdPer1M: 0.15, + }, { modelKey: 'openrouter:sao10k/l3-lunaris-8b', inputUsdPer1M: 0.04, outputUsdPer1M: 0.05 }, { modelKey: 'openrouter:sao10k/l3.1-euryale-70b', inputUsdPer1M: 0.85, outputUsdPer1M: 0.85 }, { modelKey: 'openrouter:sao10k/l3.3-euryale-70b', inputUsdPer1M: 0.65, outputUsdPer1M: 0.75 }, @@ -2798,9 +3022,9 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:tencent/hy3-preview', - inputUsdPer1M: 0.063, - outputUsdPer1M: 0.21, - cacheReadUsdPer1M: 0.021, + inputUsdPer1M: 0.18, + outputUsdPer1M: 0.6, + cacheReadUsdPer1M: 0.06, }, { modelKey: 'openrouter:thedrummer/cydonia-24b-v4.1', @@ -2818,13 +3042,13 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'openrouter:thedrummer/unslopnemo-12b', inputUsdPer1M: 0.4, outputUsdPer1M: 0.4 }, { modelKey: 'openrouter:thinkingmachines/inkling', - inputUsdPer1M: 1, + inputUsdPer1M: 0.95, outputUsdPer1M: 4.05, - cacheReadUsdPer1M: 0.17, + cacheReadUsdPer1M: 0.16, }, { modelKey: 'openrouter:thinkingmachines/inkling-small', - inputUsdPer1M: 0.5, + inputUsdPer1M: 0.45, outputUsdPer1M: 1.2, cacheReadUsdPer1M: 0.1, }, @@ -2835,6 +3059,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 0.6, cacheReadUsdPer1M: 0.015, }, + { + modelKey: 'openrouter:upstage/solar-pro4', + inputUsdPer1M: 0.03, + outputUsdPer1M: 0.12, + cacheReadUsdPer1M: 0.006, + }, { modelKey: 'openrouter:writer/palmyra-x5', inputUsdPer1M: 0.6, outputUsdPer1M: 6 }, { modelKey: 'openrouter:xiaomi/mimo-v2.5', @@ -2892,9 +3122,9 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:z-ai/glm-5', - inputUsdPer1M: 0.95, - outputUsdPer1M: 2.55, - cacheReadUsdPer1M: 0.2, + inputUsdPer1M: 0.6, + outputUsdPer1M: 1.92, + cacheReadUsdPer1M: 0.12, }, { modelKey: 'openrouter:z-ai/glm-5-turbo', @@ -2904,15 +3134,22 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'openrouter:z-ai/glm-5.1', - inputUsdPer1M: 0.952, - outputUsdPer1M: 2.992, - cacheReadUsdPer1M: 0.1768, + inputUsdPer1M: 0.966, + outputUsdPer1M: 3.036, + cacheReadUsdPer1M: 0.1794, }, { modelKey: 'openrouter:z-ai/glm-5.2', - inputUsdPer1M: 0.6958, - outputUsdPer1M: 2.1868, - cacheReadUsdPer1M: 0.12922, + inputUsdPer1M: 0.966, + outputUsdPer1M: 3.036, + cacheReadUsdPer1M: 0.1932, + }, + { modelKey: 'openrouter:z-ai/glm-5.2:free', inputUsdPer1M: 0, outputUsdPer1M: 0 }, + { + modelKey: 'openrouter:z-ai/glm-5.3', + inputUsdPer1M: 1.4, + outputUsdPer1M: 4.4, + cacheReadUsdPer1M: 0.26, }, { modelKey: 'openrouter:z-ai/glm-5v-turbo', @@ -3146,6 +3383,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 3.48, cacheReadUsdPer1M: 0.2, }, + { + modelKey: 'togetherai:deepseek-ai/DeepSeek-V4-Pro-0813', + inputUsdPer1M: 1.32, + outputUsdPer1M: 3.96, + cacheReadUsdPer1M: 0.13, + }, { modelKey: 'togetherai:essentialai/Rnj-1-Instruct', inputUsdPer1M: 0.15, outputUsdPer1M: 0.15 }, { modelKey: 'togetherai:google/gemma-3n-E4B-it', inputUsdPer1M: 0.06, outputUsdPer1M: 0.12 }, { modelKey: 'togetherai:google/gemma-4-31B-it', inputUsdPer1M: 0.39, outputUsdPer1M: 0.97 }, @@ -3373,6 +3616,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.08, cacheWriteUsdPer1M: 0.5, }, + { + modelKey: 'vercel:alibaba/qwen3.8-2.4t-a95b', + inputUsdPer1M: 2, + outputUsdPer1M: 6, + cacheReadUsdPer1M: 0.2, + }, + { + modelKey: 'vercel:alibaba/qwen3.8-27b', + inputUsdPer1M: 0.55, + outputUsdPer1M: 3.3, + cacheReadUsdPer1M: 0.11, + }, { modelKey: 'vercel:alibaba/qwen3.8-max', inputUsdPer1M: 2, @@ -3474,6 +3729,13 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.5, cacheWriteUsdPer1M: 6.25, }, + { + modelKey: 'vercel:anthropic/claude-opus-5-fast', + inputUsdPer1M: 10, + outputUsdPer1M: 50, + cacheReadUsdPer1M: 1, + cacheWriteUsdPer1M: 12.5, + }, { modelKey: 'vercel:anthropic/claude-sonnet-4', inputUsdPer1M: 3, @@ -3542,9 +3804,9 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'vercel:deepseek/deepseek-v3.2-thinking', inputUsdPer1M: 0.62, outputUsdPer1M: 1.85 }, { modelKey: 'vercel:deepseek/deepseek-v4-flash', - inputUsdPer1M: 0.2, - outputUsdPer1M: 0.4, - cacheReadUsdPer1M: 0.04, + inputUsdPer1M: 0.13, + outputUsdPer1M: 0.26, + cacheReadUsdPer1M: 0.028, }, { modelKey: 'vercel:deepseek/deepseek-v4-flash-0731', @@ -3558,6 +3820,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 3.48, cacheReadUsdPer1M: 0.14, }, + { + modelKey: 'vercel:deepseek/deepseek-v4-pro-0813', + inputUsdPer1M: 1.32, + outputUsdPer1M: 3.96, + cacheReadUsdPer1M: 0.132, + }, { modelKey: 'vercel:google/gemini-2.5-flash', inputUsdPer1M: 0.3, @@ -3632,9 +3900,15 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'vercel:google/gemini-3.6-flash', - inputUsdPer1M: 1.5, - outputUsdPer1M: 7.5, - cacheReadUsdPer1M: 0.15, + inputUsdPer1M: 0.75, + outputUsdPer1M: 3.75, + cacheReadUsdPer1M: 0.075, + }, + { + modelKey: 'vercel:google/gemini-3.7-flash', + inputUsdPer1M: 0.75, + outputUsdPer1M: 3.75, + cacheReadUsdPer1M: 0.075, }, { modelKey: 'vercel:google/gemini-omni-flash-preview', inputUsdPer1M: 1.5, outputUsdPer1M: 9 }, { @@ -3687,6 +3961,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ { modelKey: 'vercel:meta/llama-3.3-70b', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'vercel:meta/llama-4-maverick', inputUsdPer1M: 0, outputUsdPer1M: 0 }, { modelKey: 'vercel:meta/llama-4-scout', inputUsdPer1M: 0, outputUsdPer1M: 0 }, + { + modelKey: 'vercel:meta/muse-glimmer-30b', + inputUsdPer1M: 0.35, + outputUsdPer1M: 1.5, + cacheReadUsdPer1M: 0.04, + }, { modelKey: 'vercel:meta/muse-spark-1.1', inputUsdPer1M: 1.25, @@ -3831,6 +4111,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 2.4, cacheReadUsdPer1M: 0.12, }, + { + modelKey: 'vercel:nvidia/nemotron-3.5-lightning', + inputUsdPer1M: 0.05, + outputUsdPer1M: 0.2, + cacheReadUsdPer1M: 0.01, + }, { modelKey: 'vercel:nvidia/nemotron-nano-12b-v2-vl', inputUsdPer1M: 0.2, outputUsdPer1M: 0.6 }, { modelKey: 'vercel:nvidia/nemotron-nano-9b-v2', inputUsdPer1M: 0.06, outputUsdPer1M: 0.23 }, { modelKey: 'vercel:openai/gpt-3.5-turbo', inputUsdPer1M: 0.5, outputUsdPer1M: 1.5 }, @@ -3841,30 +4127,60 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 8, cacheReadUsdPer1M: 0.5, }, + { + modelKey: 'vercel:openai/gpt-4.1-fast', + inputUsdPer1M: 3.5, + outputUsdPer1M: 14, + cacheReadUsdPer1M: 0.875, + }, { modelKey: 'vercel:openai/gpt-4.1-mini', inputUsdPer1M: 0.4, outputUsdPer1M: 1.6, cacheReadUsdPer1M: 0.1, }, + { + modelKey: 'vercel:openai/gpt-4.1-mini-fast', + inputUsdPer1M: 0.7, + outputUsdPer1M: 2.8, + cacheReadUsdPer1M: 0.175, + }, { modelKey: 'vercel:openai/gpt-4.1-nano', inputUsdPer1M: 0.1, outputUsdPer1M: 0.4, cacheReadUsdPer1M: 0.025, }, + { + modelKey: 'vercel:openai/gpt-4.1-nano-fast', + inputUsdPer1M: 0.2, + outputUsdPer1M: 0.8, + cacheReadUsdPer1M: 0.05, + }, { modelKey: 'vercel:openai/gpt-4o', inputUsdPer1M: 2.5, outputUsdPer1M: 10, cacheReadUsdPer1M: 1.25, }, + { + modelKey: 'vercel:openai/gpt-4o-fast', + inputUsdPer1M: 4.25, + outputUsdPer1M: 17, + cacheReadUsdPer1M: 2.125, + }, { modelKey: 'vercel:openai/gpt-4o-mini', inputUsdPer1M: 0.15, outputUsdPer1M: 0.6, cacheReadUsdPer1M: 0.075, }, + { + modelKey: 'vercel:openai/gpt-4o-mini-fast', + inputUsdPer1M: 0.25, + outputUsdPer1M: 1, + cacheReadUsdPer1M: 0.125, + }, { modelKey: 'vercel:openai/gpt-4o-mini-search-preview', inputUsdPer1M: 0.15, @@ -3884,12 +4200,24 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 10, cacheReadUsdPer1M: 0.13, }, + { + modelKey: 'vercel:openai/gpt-5-fast', + inputUsdPer1M: 2.5, + outputUsdPer1M: 20, + cacheReadUsdPer1M: 0.25, + }, { modelKey: 'vercel:openai/gpt-5-mini', inputUsdPer1M: 0.25, outputUsdPer1M: 2, cacheReadUsdPer1M: 0.025, }, + { + modelKey: 'vercel:openai/gpt-5-mini-fast', + inputUsdPer1M: 0.45, + outputUsdPer1M: 3.6, + cacheReadUsdPer1M: 0.045, + }, { modelKey: 'vercel:openai/gpt-5-nano', inputUsdPer1M: 0.05, @@ -3915,18 +4243,18 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 2, cacheReadUsdPer1M: 0.03, }, - { - modelKey: 'vercel:openai/gpt-5.1-instant', - inputUsdPer1M: 1.25, - outputUsdPer1M: 10, - cacheReadUsdPer1M: 0.13, - }, { modelKey: 'vercel:openai/gpt-5.1-thinking', inputUsdPer1M: 1.25, outputUsdPer1M: 10, cacheReadUsdPer1M: 0.125, }, + { + modelKey: 'vercel:openai/gpt-5.1-thinking-fast', + inputUsdPer1M: 2.5, + outputUsdPer1M: 20, + cacheReadUsdPer1M: 0.25, + }, { modelKey: 'vercel:openai/gpt-5.2', inputUsdPer1M: 1.75, @@ -3939,31 +4267,49 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 14, cacheReadUsdPer1M: 0.175, }, - { modelKey: 'vercel:openai/gpt-5.2-pro', inputUsdPer1M: 21, outputUsdPer1M: 168 }, { - modelKey: 'vercel:openai/gpt-5.3-chat', - inputUsdPer1M: 1.75, - outputUsdPer1M: 14, - cacheReadUsdPer1M: 0.175, + modelKey: 'vercel:openai/gpt-5.2-fast', + inputUsdPer1M: 3.5, + outputUsdPer1M: 28, + cacheReadUsdPer1M: 0.35, }, + { modelKey: 'vercel:openai/gpt-5.2-pro', inputUsdPer1M: 21, outputUsdPer1M: 168 }, { modelKey: 'vercel:openai/gpt-5.3-codex', inputUsdPer1M: 1.75, outputUsdPer1M: 14, cacheReadUsdPer1M: 0.175, }, + { + modelKey: 'vercel:openai/gpt-5.3-codex-fast', + inputUsdPer1M: 3.5, + outputUsdPer1M: 28, + cacheReadUsdPer1M: 0.35, + }, { modelKey: 'vercel:openai/gpt-5.4', inputUsdPer1M: 2.5, outputUsdPer1M: 15, cacheReadUsdPer1M: 0.25, }, + { + modelKey: 'vercel:openai/gpt-5.4-fast', + inputUsdPer1M: 5, + outputUsdPer1M: 30, + cacheReadUsdPer1M: 0.5, + }, { modelKey: 'vercel:openai/gpt-5.4-mini', inputUsdPer1M: 0.75, outputUsdPer1M: 4.5, cacheReadUsdPer1M: 0.075, }, + { + modelKey: 'vercel:openai/gpt-5.4-mini-fast', + inputUsdPer1M: 1.5, + outputUsdPer1M: 9, + cacheReadUsdPer1M: 0.15, + }, { modelKey: 'vercel:openai/gpt-5.4-nano', inputUsdPer1M: 0.2, @@ -3977,6 +4323,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 30, cacheReadUsdPer1M: 0.5, }, + { + modelKey: 'vercel:openai/gpt-5.5-fast', + inputUsdPer1M: 12.5, + outputUsdPer1M: 75, + cacheReadUsdPer1M: 1.25, + }, { modelKey: 'vercel:openai/gpt-5.5-pro', inputUsdPer1M: 30, outputUsdPer1M: 180 }, { modelKey: 'vercel:openai/gpt-5.6-luna', @@ -3985,12 +4337,26 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.02, cacheWriteUsdPer1M: 0.25, }, + { + modelKey: 'vercel:openai/gpt-5.6-luna-fast', + inputUsdPer1M: 0.4, + outputUsdPer1M: 2.4, + cacheReadUsdPer1M: 0.04, + cacheWriteUsdPer1M: 0.25, + }, { modelKey: 'vercel:openai/gpt-5.6-sol', + inputUsdPer1M: 2.5, + outputUsdPer1M: 15, + cacheReadUsdPer1M: 0.25, + cacheWriteUsdPer1M: 3.125, + }, + { + modelKey: 'vercel:openai/gpt-5.6-sol-fast', inputUsdPer1M: 5, outputUsdPer1M: 30, cacheReadUsdPer1M: 0.5, - cacheWriteUsdPer1M: 6.25, + cacheWriteUsdPer1M: 3.125, }, { modelKey: 'vercel:openai/gpt-5.6-terra', @@ -3999,6 +4365,13 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.2, cacheWriteUsdPer1M: 2.5, }, + { + modelKey: 'vercel:openai/gpt-5.6-terra-fast', + inputUsdPer1M: 4, + outputUsdPer1M: 24, + cacheReadUsdPer1M: 0.4, + cacheWriteUsdPer1M: 2.5, + }, { modelKey: 'vercel:openai/gpt-image-1', inputUsdPer1M: 5, @@ -4063,6 +4436,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 40, cacheReadUsdPer1M: 2.5, }, + { + modelKey: 'vercel:openai/o3-fast', + inputUsdPer1M: 3.5, + outputUsdPer1M: 14, + cacheReadUsdPer1M: 0.875, + }, { modelKey: 'vercel:openai/o3-mini', inputUsdPer1M: 1.1, @@ -4076,6 +4455,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 4.4, cacheReadUsdPer1M: 0.275, }, + { + modelKey: 'vercel:openai/o4-mini-fast', + inputUsdPer1M: 2, + outputUsdPer1M: 8, + cacheReadUsdPer1M: 0.5, + }, { modelKey: 'vercel:poolside/laguna-s-2.1', inputUsdPer1M: 0.1, @@ -4089,6 +4474,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 30, cacheReadUsdPer1M: 0.5, }, + { + modelKey: 'vercel:sakana/namazu', + inputUsdPer1M: 0.95, + outputUsdPer1M: 4, + cacheReadUsdPer1M: 0.15, + }, { modelKey: 'vercel:stepfun/step-3.5-flash', inputUsdPer1M: 0.09, @@ -4174,6 +4565,7 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ cacheReadUsdPer1M: 0.2, }, { modelKey: 'vercel:xai/grok-4.5', inputUsdPer1M: 2, outputUsdPer1M: 6, cacheReadUsdPer1M: 0.3 }, + { modelKey: 'vercel:xai/grok-4.6', inputUsdPer1M: 2, outputUsdPer1M: 6, cacheReadUsdPer1M: 0.5 }, { modelKey: 'vercel:xai/grok-build-0.1', inputUsdPer1M: 1, @@ -4216,12 +4608,6 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 2.2, cacheReadUsdPer1M: 0.11, }, - { - modelKey: 'vercel:zai/glm-4.6v', - inputUsdPer1M: 0.3, - outputUsdPer1M: 0.9, - cacheReadUsdPer1M: 0.05, - }, { modelKey: 'vercel:zai/glm-4.7', inputUsdPer1M: 0.6, @@ -4250,9 +4636,9 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ }, { modelKey: 'vercel:zai/glm-5.2', - inputUsdPer1M: 1.1, - outputUsdPer1M: 3.851, - cacheReadUsdPer1M: 0.275, + inputUsdPer1M: 0.8, + outputUsdPer1M: 2.55, + cacheReadUsdPer1M: 0.16, }, { modelKey: 'vercel:zai/glm-5.2-fast', @@ -4260,6 +4646,12 @@ export const GENERATED_MODEL_PRICING: readonly PricingConfig[] = [ outputUsdPer1M: 6.6, cacheReadUsdPer1M: 0.21, }, + { + modelKey: 'vercel:zai/glm-5.3', + inputUsdPer1M: 1.4, + outputUsdPer1M: 4.4, + cacheReadUsdPer1M: 0.26, + }, { modelKey: 'vercel:zai/glm-5v-turbo', inputUsdPer1M: 1.2, diff --git a/scripts/sync-model-metadata.mjs b/scripts/sync-model-metadata.mjs index 0843ed66bd..f86876af73 100644 --- a/scripts/sync-model-metadata.mjs +++ b/scripts/sync-model-metadata.mjs @@ -264,6 +264,7 @@ export function toMetadata(providerId, modelId, provider, model) { maxOutputTokens: model.limit?.output, ...(model.knowledge !== undefined ? { knowledgeCutoff: model.knowledge } : {}), ...(model.structured_output !== undefined ? { structuredOutput: model.structured_output } : {}), + ...(model.cost?.input === 0 && model.cost?.output === 0 ? { isFree: true } : {}), ...(model.last_updated !== undefined ? { lastUpdated: model.last_updated } : {}), capabilities: { ...(model.modalities ? { vision: model.modalities.input.includes('image') } : {}),