minimax/minimax-m2
MiniMax M2.
Legacy chat model, kept for clients pinned to it.
| Slug | minimax/minimax-m2 |
| Kind | chat |
| Vendor | minimax |
| Endpoint | POST /v1/chat/completions |
Charged per job from what the generator reports it rendered — your Logs show the exact figure for each one; see balance and billing.
Calling it
curl https://api.sociaro.com/v1/chat/completions \
-H "Authorization: Bearer $SOCIARO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "minimax/minimax-m2",
"messages": [{"role": "user", "content": "hello"}]
}'Parameters
| Field | Type | Required | Default | Values | What it does |
|---|---|---|---|---|---|
messages | object[] | yes | — | OpenAI chat messages (system, user, assistant, tool) | The conversation, in OpenAI's format. |
max_completion_tokens | integer | no | the model's | 1 up to the model's output limit | Thinking tokens count toward it: too small a cap ends with `finish_reason: length` and an empty answer. |
max_tokens | integer | no | — | as above | The older name for the same cap; MiniMax marks it deprecated. |
stream | boolean | no | false | true · false | Server-sent events, as in OpenAI's API. |
stream_options | object | no | — | {"include_usage": true} | Adds a final chunk with token usage. |
temperature | number | no | 1 | (0, 1] | MiniMax documents it as having no effect on M3 (fixed at 1). |
top_p | number | no | 0.95 | (0, 1] | MiniMax documents it as having no effect on M3 (fixed at 0.95). |
tools | object[] | no | — | OpenAI function tools | Function calling, in OpenAI's format. |
thinking | object | no | on | {"type": "adaptive"} | These models always reason before answering; reasoning cannot be turned off. |
reasoning_split | boolean | no | the vendor's | true · false | true returns the reasoning in `reasoning_content`; false leaves it inside the answer between think tags. |