Skip to content

chore: update model catalog from bot issues#909

Merged
Erin McNulty (erin2722) merged 2 commits into
mainfrom
chore/autofix-bot-issues-2026-07-03
Jul 6, 2026
Merged

chore: update model catalog from bot issues#909
Erin McNulty (erin2722) merged 2 commits into
mainfrom
chore/autofix-bot-issues-2026-07-03

Conversation

@github-actions

@github-actions github-actions Bot commented Jul 3, 2026

Copy link
Copy Markdown
Contributor

Automated daily batch of model catalog updates from bot issues.

Included issues

Summary

Issue Provider Primary model Changed models Added models Updated models Verification sources
#907 bedrock us.anthropic.claude-sonnet-5 us.anthropic.claude-sonnet-5
global.anthropic.claude-sonnet-5
us.anthropic.claude-sonnet-5
global.anthropic.claude-sonnet-5
None 1
2
#908 google gemini-3.1-flash-lite-image gemini-3.1-flash-lite-image
publishers/google/models/gemini-3.1-flash-lite-image
gemini-3.1-flash-lite-image
publishers/google/models/gemini-3.1-flash-lite-image
None 1
2
3

Verified metadata

#907: [BOT ISSUE] Bedrock: add missing us.anthropic.claude-sonnet-5 and global.anthropic.claude-sonnet-5 entries

Model Display name Parent Providers Format Flavor Token limits Pricing Lifecycle
us.anthropic.claude-sonnet-5 US Claude Sonnet 5 anthropic.claude-sonnet-5 bedrock anthropic chat input=1000000, output=128000 n/a parent=anthropic.claude-sonnet-5; multimodal=true; reasoning=true
global.anthropic.claude-sonnet-5 Global Claude Sonnet 5 anthropic.claude-sonnet-5 bedrock anthropic chat input=1000000, output=128000 n/a parent=anthropic.claude-sonnet-5; multimodal=true; reasoning=true

Verification notes

No LLM verification step ran; model metadata was already complete in the issue.

sync_models vs proposed update

sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.

Model Field Proposed update sync_models sync_models source models
us.anthropic.claude-sonnet-5 input_cost_per_mil_tokens n/a 3.3 us.anthropic.claude-sonnet-5
us.anthropic.claude-sonnet-5 output_cost_per_mil_tokens n/a 16.5 us.anthropic.claude-sonnet-5
us.anthropic.claude-sonnet-5 input_cache_read_cost_per_mil_tokens n/a 0.33 us.anthropic.claude-sonnet-5
us.anthropic.claude-sonnet-5 input_cache_write_cost_per_mil_tokens n/a 4.125 us.anthropic.claude-sonnet-5
global.anthropic.claude-sonnet-5 input_cost_per_mil_tokens n/a 3 global.anthropic.claude-sonnet-5
global.anthropic.claude-sonnet-5 output_cost_per_mil_tokens n/a 15 global.anthropic.claude-sonnet-5
global.anthropic.claude-sonnet-5 input_cache_read_cost_per_mil_tokens n/a 0.3 global.anthropic.claude-sonnet-5
global.anthropic.claude-sonnet-5 input_cache_write_cost_per_mil_tokens n/a 3.75 global.anthropic.claude-sonnet-5

#908: [BOT ISSUE] Google: add missing gemini-3.1-flash-lite-image model

Model Display name Parent Providers Format Flavor Token limits Pricing Lifecycle
gemini-3.1-flash-lite-image Gemini 3.1 Flash-Lite Image google, vertex google chat input=65536, output=4096 in/out=0.25/1.5 per 1M multimodal=true
publishers/google/models/gemini-3.1-flash-lite-image Gemini 3.1 Flash-Lite Image vertex google chat input=65536, output=4096 in/out=0.25/1.5 per 1M multimodal=true

Verification notes

No LLM verification step ran; model metadata was already complete in the issue.

sync_models vs proposed update

sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.

Model Field Proposed update sync_models sync_models source models
gemini-3.1-flash-lite-image catalog entry present missing None
publishers/google/models/gemini-3.1-flash-lite-image catalog entry present missing None

@vercel

vercel Bot commented Jul 3, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
ai-proxy Ready Ready Preview, Comment Jul 6, 2026 2:18pm

Request Review

@github-actions

github-actions Bot commented Jul 3, 2026

Copy link
Copy Markdown
Contributor Author

Codex (@codex) review

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: b4a40af348

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "Codex (@codex) review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "Codex (@codex) address that feedback".

"grok-latest": ["xAI"],
"gemma-4-31b": ["cerebras"],
"gpt-5.4-mini-2026-03-17": ["openai", "azure"],
"gemini-3.1-flash-lite-image": ["google", "vertex"],

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Keep the short Gemini ID Google-only

For this exact short model ID, adding vertex makes getDirectModelEndpointTypes("gemini-3.1-flash-lite-image") route Vertex credentials directly to fetchGoogleChatCompletions; that path appends the raw model string into the Vertex URL (.../locations/<loc>/${model}:generateContent), so a Vertex call uses .../gemini-3.1-flash-lite-image instead of the publisher resource added on the next line. The existing BT-5895 test in packages/proxy/schema/index.test.ts documents that short Gemini IDs should remain Google-only and get Vertex via the fallback publisher model.

Useful? React with 👍 / 👎.

"bedrock"
]
},
"global.anthropic.claude-sonnet-5": {

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Restore Bedrock pricing for Sonnet 5 regional IDs

These new Bedrock Anthropic IDs are the only Bedrock Claude entries without input_cost_per_mil_tokens/output_cost_per_mil_tokens and cache price fields, and catalog consumers do not inherit prices through parent; when callers use global.anthropic.claude-sonnet-5 or the new us.anthropic.claude-sonnet-5 ID directly, cost reporting will be blank/zero despite AWS publishing nonzero Sonnet 5 Bedrock rates. Please carry the resolved global vs US regional pricing fields onto the new entries.

Useful? React with 👍 / 👎.

#907 added us.anthropic.claude-sonnet-5 and global.anthropic.claude-sonnet-5 but
without input/output pricing, unlike every other Bedrock Claude entry (Codex
P2). Mirror anthropic.claude-sonnet-5: $3 / $15 per MTok plus cache read/write
rates, so cost tracking works for callers using the regional ids directly.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@erin2722 Erin McNulty (erin2722) merged commit 697e5f7 into main Jul 6, 2026
5 of 6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

1 participant