You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
MiniMax-M2 reasoning content not properly handled in OpenAI-compatible mode, should we support this ? #3555
After testing, I discovered the issue. The API returns reasoning content wrapped in think tags within delta.content instead of placing it in delta.reasoning_content:
As you can see, the reasoning content is wrapped in "think" tags in delta.content rather than being placed in delta.reasoning_content.
I can fix this by adding functions to handle the message processing, but should this be necessary? Does this behavior align with the OpenAI API specification?
MiniMax-M2 is an 'interleaved thinking model', which requires the thinking content to be added to the context. See the model card on HuggingFace: https://huggingface.co/MiniMaxAI/MiniMax-M2.
MiniMax-M2 is an 'interleaved thinking model', which requires the thinking content to be added to the context. See the model card on HuggingFace: https://huggingface.co/MiniMaxAI/MiniMax-M2.
Activity
rekram1-node commentedon Oct 30, 2025
if reasoning parts are returned in content chunks instead of actual reasoning chunks then it isnt following the spec
whoisjeremylam commentedon Oct 30, 2025
MiniMax-M2 is an 'interleaved thinking model', which requires the thinking content to be added to the context. See the model card on HuggingFace: https://huggingface.co/MiniMaxAI/MiniMax-M2.
rekram1-node commentedon Oct 30, 2025
i think the ai sdk may provide some utility to automatically parse the reasoning chunjs for these types of models
zerogog commentedon Oct 30, 2025
Good to know this.
pascalandy commentedon Oct 30, 2025
Works with this config:
but it gets 10K context window from a simple ping:
Add community research on MiniMax-M2 latency to opencode-analysis.md