Mistral Large 3, not served here yet
Mistral Large 3 is a 675B mixture-of-experts with 41B active parameters, trained from scratch and released under Apache 2.0. OpenRelay does not serve it yet; tell us if you want it.
- Text in, text out
- Not served yet
- No public rate
Not served yet.
GET /v1/models lists Mistral Large 3 with status request_access, but nothing serves it: there is no backend behind the id and no public rate. A request tells us someone wants it.
openrelay/mistral-large-3Tell us you want Mistral Large 3
Leave a work email. We email you if we start serving it, and nothing else goes to this list.
The spec sheet.
What Mistral AI has published about the model, with the source on each row.
- On OpenRelay
- openrelay/mistral-large-3: Text in, text out. Listed with status request_access, not served. GET /v1/models
- Parameters
- 675B total, 41B active (granular mixture-of-experts) Mistral AI model card
- Context
- 256K tokens Mistral AI model card
- Tool use
- Native function calling and JSON output Mistral AI model card
- Limits, per Mistral
- Not a dedicated reasoning model: dedicated reasoning models can outperform it on strict reasoning Mistral AI model card
- License
- Apache 2.0, open weights Mistral AI, December 2, 2025
- Released
- December 2, 2025 Mistral AI, December 2, 2025
Mistral Large 3, answered.
Can I call Mistral Large 3 through OpenRelay?
Not today. openrelay/mistral-large-3 appears on GET /v1/models with status request_access, but nothing serves it: a chat completions call returns 403 model_access_required. The form on this page tells us you want it.
Is Mistral Large 3 a reasoning model?
Not a dedicated one, in Mistral's own words. Its model card points it at instruction following, tool use and long documents, and says dedicated reasoning models can beat it on strict reasoning.