Mistral: Mistral Medium 3

Mistral: Mistral Medium 3

mistralai · Released May 7, 2025
52
Our Score

Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases. The model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.

$0.40 / 1M Input Price
$2.00 / 1M Output Price
131,072 tokens Context Window

Capabilities

Tool Use Function Calling Vision

Architecture

ModalityText + Image → Text
TokenizerMistral

Performance Indices

Source: Artificial Analysis

18.8 Intelligence Index
13.6 Coding Index
14.1 Agentic Index
30.3 Math Index

Benchmark Scores

Evaluations

GPQA Diamond 57.8%
Graduate-level scientific reasoning
HLE 4.3%
Humanity's Last Exam
MMLU Pro 76%
Multi-task language understanding
LiveCodeBench 40%
Live coding evaluation
SciCode 33.1%
Scientific computing
MATH 500 90.7%
Mathematical problem-solving
AIME 44%
Competition mathematics
AIME 2025 30.3%
Competition mathematics (2025)
IFBench 39.3%
Instruction following
LCR 28%
Long-context reasoning
TerminalBench Hard 3.8%
Agentic terminal tasks
τ²-Bench 24.3%
Conversational agent benchmark

Benchmark data from Artificial Analysis and Hugging Face

Model Information

OpenRouter ID mistralai/mistral-medium-3
Providermistralai
Model FamilyMistral
Release Date May 7, 2025
Context Length131,072 tokens
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.40 $0.000400
Output $2.00 $0.002000

Live Performance

Live endpoint metrics — refreshed every 30 minutes.

650ms
Best Latency (TTFT)
52 tok/s
Best Throughput
0/1
Active Endpoints
Available via: Mistral

Leaderboard Categories