modelgrep
M

Mistral: Mistral Medium 3

mistralai/mistral-medium-3

123rd smartest of 179ToolsJSONVision
Use via OpenRouter ↗
Intelligence
18.8
123rd of 179
Design Elo
1163
3D
Speed
35
231st fastest
Latency
446ms
first token
Input price
$0.400
150th cheapest
Context
131K

How it compares

Smarter than31%
of all ranked models
Faster than26%
of all ranked models
Cheaper than50%
of all ranked models

Overview

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...

Benchmarks

independent · via OpenRouter
Artificial Analysis30th percentile
Intelligence Index
18.8
Coding Index
13.6
Agentic Index
13.7
GPQA Diamond
58%
Humanity's Last Exam
4%
SciCode
33%
Tau²-Bench (agentic)
24%
Design Arena · Elo2,890 tournaments
3D
1163
Website
1124
codecategories
1121
Game Dev
1087
UI Component
1084
Data Viz
1076

Providers & pricing (1)

ProviderIn $/MOut $/MUptime
Mistral$0.400$2.00100%

Specifications

Context window131K
Max output
Knowledge cutoffMar 2025
Input modalitiestext, image, file
Output modalitiestext
Prompt caching
Cache read price$0.040/M
ModeratedNo

Mistral Medium 3 FAQ

How much does Mistral Medium 3 cost?

Mistral Medium 3 costs $0.400 per million input tokens and $2.00 per million output tokens via OpenRouter, making it 150th cheapest of 298 paid models.

How smart is Mistral Medium 3?

Mistral Medium 3 scores 18.8 on the Artificial Analysis Intelligence Index, ranking 123rd of 179 benchmarked models, with a GPQA Diamond score of 58%.

How fast is Mistral Medium 3?

Mistral Medium 3 generates around 35 tokens per second with 446ms time-to-first-token (p50), the 231st fastest tracked model.

What is Mistral Medium 3's context window?

Mistral Medium 3 supports a 131K-token context window. It accepts text, image, file input.

Compare head-to-head