Back to all models
M

Mistral Large

MistralReleased 2024-06-06
Official Page

Mistral AI's most powerful model with excellent multilingual capabilities, strong coding, and competitive European AI development.

Overall Score
76%
Cost Efficiency
100
Context Window
128K

Strengths

  • ✓Strong multilingual support
  • ✓European AI sovereignty
  • ✓Good value pricing
  • ✓Vision capable

Weaknesses

  • ✗Lagging behind top models
  • ✗Smaller context window
Architecture
Transformer (MoE)
Parameters
129B (12B active)
Input
$2/M
Output
$6/M

Benchmark Performance

BenchmarkDescriptionScoreCategoryBar
MMLUMassive Multitask Language Understanding84.9%knowledge
HumanEvalCode generation accuracy87.7%coding
MATHMathematical reasoning83.5%math
GSM8KGrade school math88.2%math
GPQAGraduate-level science Q&A49.1%reasoning
HellaSwagCommon sense reasoning85.7%reasoning
SWE-benchSoftware engineering problems34.9%coding
ARC-CAI2 Reasoning Challenge93%reasoning
AI
ModelBench

The open benchmark platform for comparing the latest AI models and their capabilities.

Platform

  • Compare Models
  • All Models
  • Methodology

AI Labs

  • OpenAI
  • Anthropic
  • Google DeepMind
  • Meta AI

Resources

  • Documentation
  • Benchmark Guide
  • API Reference
Benchmarks are sourced from official model cards and independent evaluations. Data is for informational purposes only.
© 2026 ModelBench. Built for the AI community.