Mistral Large 2.0
Mistral Large 2.0 introduced the 123B Large model with an advertised 128K context. It supports multilingual text and coding, with downloadable instruction weights under research terms.
A dense model for long documents
The July launch describes 123B parameters and support for dozens of languages and more than 80 programming languages. It also emphasizes function calling, retrieval and following precise instructions across a conversation. The parameter count is dense, not an active subset of a larger expert model.Mistral Large 2.0 release announcement
Hosted retirement and research weights
The versioned API mistral-large-2407 is retired. The current catalog still links Mistral-Large-Instruct-2407 weights under MRL. Commercial self-deployment requires separate terms from Mistral. This access distinction matters when reproducing an older Large 2 application.Mistral Large 2.0 model and lifecycle recordMistral Large 2.0 release announcement
Specifications & access
| Detail | Value and conditions |
|---|---|
| Launch date | |
| Access | Downloadable |
| Model identifiers | Mistral-Large-Instruct-2407mistral-large-2407 |
| Context window | 128,000 tokens The model catalog advertises 128k context for this version. This is not a measured recall limit. Mistral Large 2.0 model and lifecycle record |
| Weight license | Mistral Research License; commercial use needs separate permission/licenseMistral Large 2.0 release announcement |
| Access and lifecycle | Hosted API retired on 30 March 2025. Downloadable weights remain listed.Mistral Large 2.0 model and lifecycle recordMistral Large 2.0 release announcement |
| Parameters | 123 billionMistral Large 2.0 model and lifecycle record |