Mistral releases Mistral Large 2 with 128K context and improved multilingual support

Mistral AI has released Mistral Large 2, the latest iteration of its flagship model, which offers significant advancements in code generation, mathematics, reasoning, and multilingual support. The model features a 128k context window, 123 billion parameters optimized for single-node inference, and enhanced function-calling capabilities.

Achieves 84.0% on the MMLU benchmark, outperforming its predecessor and competing with models like GPT-4o and Claude 3 Opus.
Supports over 80 coding languages and excels in English, French, German, Spanish, Italian, Portuguese, Dutch, Russian, Chinese, Japanese, Korean, Arabic, and Hindi.
Available on "la Plateforme" under the API name mistral-large-2407 and hosted on HuggingFace, with commercial use requiring a Mistral Commercial License.
Expanded availability through partnerships with Google Cloud Platform, Azure AI Studio, Amazon Bedrock, and IBM watsonx.ai.

The release consolidates Mistral's offerings around general-purpose and specialist models, providing users with high-performing, cost-efficient options for complex business applications and multilingual tasks.

Benchmarks

Benchmark	Model	Score
MMLU	Mistral Large 2	84%

Benchmark

Model

Score

MMLU

Mistral Large 2

84%

Benchmarks

Mistral launches regional endpoints, open model support, and European compute coalition

mistral.rs v0.8.10 adds /v1/skills support for local models

Mistral releases Large 2 with 123B parameters for cost-efficient single-node inference

Mistral AI releases Shieldstral 1.0 3B, a policy-adaptive multimodal safety classifier

Mistral releases Shieldstral-1.0-3B safety model