feat(metamodels): Add support for Muse Spark 1.1 via the Meta Models API - #10432
Conversation
`ThinkingEffort::Off` is not supported by the Muse Spark models, so I've clamped it to `ThinkingEffort::low` which most closely matches the user's intent to have no reasoning. I'd love to get feedback on whether this approach makes sense, or if there's a better way to do it.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 14c1707222
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: d45bbe423f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| "reasoning": true | ||
| } | ||
| ], | ||
| "preserves_thinking": true, |
There was a problem hiding this comment.
This provider still routes muse-spark-1.1 through the chat-completions path, but setting preserves_thinking makes goose echo any captured reasoning_content from prior assistant turns back into later requests. Meta's reasoning docs distinguish replayable encrypted reasoning on the Responses API from chat-completions reasoning_content, so multi-turn/tool sessions can bloat prompts with large thinking traces or rely on continuity Meta does not provide; set this false unless the provider is routed through /v1/responses.
Useful? React with 👍 / 👎.
|
It appears that |
|
Works great otherwise, I tried it with 
|
Thanks for finding this. I've removed the commit adding support for "none." |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 14c1707222
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
The-Best-Codes
left a comment
The-Best-Codes
left a comment
There was a problem hiding this comment.
Looks great! Working locally.
3 participants
Summary
Add support for the Meta Models API and the Muse Spark 1.1 model.
NB: Replaces #10399, which I messed up.
Testing
goose configureIf someone with access to the model can test, we can merge this, or we could wait until I get access.
Related Issues
#10396
Screenshots/Demos (for UX changes)