diff --git a/docs/config_example/README.md b/docs/config_example/README.md index 5284b471..42558b7b 100644 --- a/docs/config_example/README.md +++ b/docs/config_example/README.md @@ -35,6 +35,7 @@ Do not paste secrets into public issues or pull requests. Configure provider cre - [A2Agent](a2agent.md) - [AI-ROUTER](ai-router.md) - [DeepSeek](deepseek.md) +- [UberLLM](uberllm.md) ## Community Example PR Scope diff --git a/docs/config_example/uberllm.md b/docs/config_example/uberllm.md new file mode 100644 index 00000000..78c46aa9 --- /dev/null +++ b/docs/config_example/uberllm.md @@ -0,0 +1,40 @@ +# UberLLM Configuration Example + +This example uses the primary OpenCode v2 configuration shape. For legacy OpenCode v1, adapt using the [migration guide](../migrate-v1-to-v2.md). + +## Disclaimer + +This project is not affiliated with, endorsed by, or sponsored by [UberLLM](https://uberllm.dev/). + +## Example + +```jsonc +{ + "$schema": "https://opencode.ai/config.json", + "plugins": [ + "opencode-models-discovery@2.1.0" + ], + "providers": { + "uberllm": { + "package": "@opencode/ai/providers/openai-compatible", + "settings": { + "baseURL": "https://api.uberllm.dev/v1", + // Optional: omit if managing credentials interactively via OpenCode '/connect' + "apiKey": "example_api_key", + "modelsDiscovery": { + "enabled": true, + "modelInfoFormat": "openrouter" + } + } + } + } +} +``` + +## Notes + +- Configure credentials with OpenCode `/connect` when possible, or through local private configuration. +- This uses the plugin's standard `/v1/models` discovery path, which resolves to `https://api.uberllm.dev/v1/models`. That endpoint lists chat models only; image, video and music models are not returned there. +- The `/v1/models` response carries OpenRouter-style inline fields (`context_length`, `top_provider.max_completion_tokens`, `architecture` modalities, `supported_parameters`, `pricing`), so `modelInfoFormat: "openrouter"` fills limits, modalities, tool-call support and cost from the same response without another request. Remove it if you do not want the plugin to set per-model costs. +- The example intentionally does not declare a static model list or filters. Returned models depend on the current API response. +- Verify the current API base URL, model availability, pricing, terms of service, data handling, and regional availability before use.