Ai-Bitz
Back to Models

Mistral Large 2

Best for Tools
Mistral
Released: 2024-07-24
Type: LLM
Context
128k
Max Out
4.1k
Cutoff
2024-06
Params
123B

The Breakdown

Mistral Large 2 represents the pinnacle of European AI development. It focuses heavily on 'reasoning density'—delivering high intelligence in a relatively compact 123B parameter frame. It is particularly renowned for its function-calling (tool use) capabilities and strict adherence to JSON schemas, making it a favorite for building reliable agentic workflows. Unlike US-centric models, it boasts superior fluency in French, German, Spanish, and Italian, positioning it as the go-to choice for global enterprise deployments.

Overall Score
9
/10
Pricing (per 1k tokens)
Input$0.003
Output$0.009
Currency: USD

Version History

Mixtral 8x7B
2023-12-11
9.1/10
Mistral 7B
2023-09-27
8.9/10

The Good

  • Exceptional function calling and JSON mode capabilities
  • Strong multilingual support (European languages)
  • Can be fine-tuned or hosted privately

The Bad

  • Smaller ecosystem than OpenAI/Google
  • Slightly expensive compared to Llama 3 70B
  • General knowledge slightly lags behind GPT-4o

The Verdict

The European champion. If you need strict data sovereignty or precise tool use without the 'black box' vibes of OpenAI, Mistral Large 2 is a serious contender. It punches above its weight in reasoning.

Performance Benchmarks

drop
81
human eval
86.5
mmlu
84