Mistral
Released August 12, 2025Cutoff June 2025

Mistral Medium 3.1

Mistral Medium 3.1 (mistral-medium-2508) by Mistral AI: frontier multimodal model with 128k context, function calling, and enterprise adaptation support.

Inputs
Text
Image
File
Outputs
Text

Model Overview

Capabilities, design details, and architectural traits

Mistral Medium 3.1 - Frontier-Class Multimodal Enterprise Model

Mistral Medium 3.1 (mistral-medium-2508) is Mistral AI's frontier-class multimodal model, officially described as improving tone and performance over its predecessor. It is classified as a Premier tier model - a commercial-only offering within Mistral's lineup - and has since been superseded by Mistral Medium 3.5.

Feature Set

Officially documented API capabilities include: Chat Completions, Function Calling, Agents & Conversations, Built-In Tools, Structured Outputs, Predicted Outputs, Prefix, OCR (with Annotations and BBox Extraction), Document QnA, FIM (Fill-in-the-Middle), Embeddings, Moderations, Chat Moderations, Transcriptions, Text to Speech, Timestamps, and Batching.

Deployment & Enterprise Design

Mistral AI documents Mistral Medium 3.1 as supporting hybrid, on-premises, and in-VPC deployment, as well as continuous pretraining, full fine-tuning, and integration into enterprise knowledge bases - enabling domain-specific adaptation without starting from scratch.

Why It Exists

Mistral Medium 3.1 was positioned as a frontier-quality model delivering enterprise adaptation capabilities and multimodal understanding, continuing the Medium family's design goal of balancing performance and deployability.

Benchmark Performance

Independent evaluations · Artificial Analysis

9.2%
Intelligence
20.5%
Coding Index
1.9%
Agentic Index

Accuracy & Capability Details

GPQA - Graduate Science58.8%
Humanity's Last Exam4.7%
SciCode - Scientific Coding32.4%
Instruction Following39.8%
Long Context Reasoning42.7%
τ²-Bench - Agentic Tasks40.6%
TerminalBench - System Control10.6%

Compare Models Side-by-Side

Evaluate specifications, pricing, and independent benchmark indices

Model Details
General Info
ProviderMistralAnthropicAnthropic
Release DateAugust 12, 2025September 22, 2026September 28, 2026
Knowledge CutoffJun 2025-Jun 2026
Context & Limits
Context Window131K
1M
Best Context Window
1M
Best Context Window
Pricing (per 1M tokens)
Input Pricing
$0.40
Best Input Pricing
$4$2
Output Pricing
$2
Best Output Pricing
$20$10
Modalities
Inputs
textimagefile
textimagefile
textimagefile
Outputs
text
text
text
Benchmarks (0-100)
Intelligence Index9.2
57.6
Best Intelligence Index
56.0
Coding Index20.5--
Agentic Index1.9--
Mistral Medium 3.1
Claude Opus 5.5
Claude Sonnet 5.5

Humanity's Last Exam

Extremely difficult logical reasoning and knowledge.

5%
Mistral Medium 3.1
Humanity's Last Exam
Score: 5%
Mistral Medium 3.1
61%
Claude Opus 5.5
Humanity's Last Exam
Score: 61%
Claude Opus 5.5
55%
Claude Sonnet 5.5
Humanity's Last Exam
Score: 55%
Claude Sonnet 5.5

Long Context Reasoning

Logical reasoning over long context windows.

43%
Mistral Medium 3.1
Long Context Reasoning
Score: 43%
Mistral Medium 3.1
85%
Claude Opus 5.5
Long Context Reasoning
Score: 85%
Claude Opus 5.5
83%
Claude Sonnet 5.5
Long Context Reasoning
Score: 83%
Claude Sonnet 5.5

SciCode Benchmark

Scientific coding and mathematical modeling.

32%
Mistral Medium 3.1
SciCode Benchmark
Score: 32%
Mistral Medium 3.1
67%
Claude Opus 5.5
SciCode Benchmark
Score: 67%
Claude Opus 5.5
61%
Claude Sonnet 5.5
SciCode Benchmark
Score: 61%
Claude Sonnet 5.5

Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.

Explore more from Mistral

Other models by Mistral

Top AI Models

Leading alternatives by intelligence score

View all