Enjoying AI Stats?Support us
FR
Mistral

Mistral Small 3.2

Mistral
Overview

No description provided. Want to help? Contribute on GitHub or click here!

Quick Links
Key Metrics
Max Input
128,000
Tokens
Max Output
128,000
Tokens
Throughput
-
tok/s
Latency
-
ms
Input Price
-
Per 1M Tokens
Cached Input Price
-
Per 1M Tokens
Output Price
-
Per 1M Tokens
Blended Price
-
Per 1M Tokens
Model Information
Release Details

Released

20 Jun 2025

Knowledge Cutoff

-

License

Apache 2.0

Model Architecture

Parameters

-

Training Data

-

Context Window

Input Context Length

128,000 tokens

Output Context Length

128,000 tokens

Key Features
Web Access
No

Real-time access to current web information

Multimodal
Unknown

Ability to process multiple data types (text, images, etc.)

Reasoning
No

Advanced logical and deductive reasoning capabilities

Fine-Tunable
Unknown

Can be customized for specific use cases

Model Release & Updates
20 Jun 2025
Model Released
Model first made available to the public
20 Jun 2025
Model Announced
Model first introduced to the public
17 Mar 2025
Previous version
Benchmarks & Performance Comparison
AI Stats Score
1394.80
#139/191
AI2D
92.90%
#1/1
ChartQA
87.40%
#1/1
DocVQA
94.90%
#1/1
GPQA
44.20%
#1/1
GPQA Diamond
46.10%
#60/94
HumanEval
92.90%
#1/5
MATH
69.40%
#1/4
Mathvista
67.10%
#1/1
MMLU
80.50%
#8/11
MMLU-Pro
69.10%
#6/6
MMMU
62.50%
#12/13
SimpleQA
12.10%
#11/11