OpenCode Go - #200
Conversation
|
All contributors have signed the CLA ✍️ ✅ |
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 5cb8c91. Configure here.
| rates(0.14, 0.28, 0.0028, 0.0), | ||
| none_and_minimal_through_xhigh_levels(), | ||
| false, | ||
| ), |
There was a problem hiding this comment.
Wrong wire API for DeepSeek Flash
High Severity
deepseek-v4-flash is seeded as OpenAiResponses, so turns post to /v1/responses. OpenCode Go currently serves that model only on Chat Completions; Responses calls fail upstream. Selecting this model makes inference fail even with a valid OPENCODE_API_KEY.
Additional Locations (1)
Reviewed by Cursor Bugbot for commit 5cb8c91. Configure here.
There was a problem hiding this comment.
Docs table is stale. Go serves flash on /responses, not /chat/completions. oh-my-pi 17.2.12 and current catalog api=openai-responses. Live /responses 400s are DeepSeek reasoning replay, not a missing-model miss.
|
I have read the CLA Document and I hereby sign the CLA |
1 participant
No description provided.