Repository navigation
[Docs] Document the Use Responses API option for OpenAI Compatible providers - #72
zoomote2[bot] wants to merge 1 commit into
Conversation
…patible Explain when to enable the option (endpoints that require /v1/responses to combine reasoning with tool use, such as Azure OpenAI deployments of GPT-6.1 Terra) and add a troubleshooting entry for the 400 reasoning_effort error.
|
2 documentation accuracy issues outstanding. See session
Reviewed 6843e25 |
|
|
||
| ## Responses API | ||
|
|
||
| By default, the OpenAI Compatible provider sends requests to the Chat Completions endpoint (`/chat/completions`). Enable **Use Responses API** in the provider settings to send requests to the Responses endpoint (`/responses`) instead. |
There was a problem hiding this comment.
This overstates the setting's scope. The paired implementation only routes createMessage() through Responses; completePrompt() still calls Chat Completions for prompt enhancement and task-title generation. Please qualify this as task/conversation requests rather than all provider requests.
| To use function tools, use /v1/responses or set reasoning_effort to 'none'. | ||
| ``` | ||
|
|
||
| With **Use Responses API** enabled: |
There was a problem hiding this comment.
This mode also needs an Extra Body caveat or corresponding implementation support. The paired handleResponsesApiMessage() never applies withExtraBody(), but this page says Extra Body fields are added to streaming requests. Users whose gateway requires those fields would otherwise expect them to be sent with Responses API requests.
What changed
The OpenAI Compatible provider page gains a Responses API section. It explains the new Use Responses API setting: requests go to
<Base URL>/responsesinstead of/chat/completions, the configured reasoning effort is sent asreasoning: { effort: ... }, and reasoning works together with native tool calling. The troubleshooting section gains an entry for the 400 "Function tools with reasoning_effort are not supported" error that points to the new section.Why this change was made
The main Zoo Code repository adds the Use Responses API setting in a paired PR. Endpoints such as Azure OpenAI deployments of GPT-6.1 Terra behind a corporate gateway require the Responses API to combine reasoning with tool use. Users need documentation for when and why to enable the option.
Impact
Documentation-only change. Readers of the OpenAI Compatible provider page can now discover the setting and self-serve the fix for the 400 error. No product behavior changes.
Screenshots
The setting this page documents (from the paired Zoo Code PR):
Related PRs
Validated with
pnpm run check-typesin this repository.