Ministral 3 8B

Ministral 3 8B

Mistral · Released Dec 2, 2025 Professional
Intelligence #10 / 576
82.0 Our Score
Speed #137 / 271
93.8 tokens / sec
Input #248 / 577
$0.150 per 1M tokens
Output #162 / 577
$0.150 per 1M tokens
Context
— Not reported

Analysis Summary

Ministral 3 8B is a small-footprint model from Mistral, positioned at the budget end of the market. Its intelligence and coding benchmarks are limited relative to the broader field, but its MMLU-Pro score of 0.642 and GPQA of 0.471 suggest it can handle general knowledge tasks and basic instruction following at a functional level.

For businesses, this model is best suited to high-volume, low-complexity tasks: content drafting, simple summarisation, or classification pipelines where cost per token matters more than depth of reasoning. Coding and agentic reliability are weak, so it is not a fit for software engineering or autonomous workflows.

At $0.15/1M tokens for both input and output, it is one of the more affordable options from a reputable Western provider. Teams running bulk content operations on a tight budget will find it serviceable, but should step up to a larger model for anything requiring nuanced reasoning or reliable instruction following.

Assessed June 6, 2026

Editorial notes

Ministral 3 8B is a compact Mistral model with modest reasoning and coding capability, suited to lightweight content tasks at a very low price point of $0.15/1M tokens.

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

Performance Profile

Intelligence2.7Technical1.7Value7Content3
Intelligence 2.7/10
Technical 1.7/10
Content 3/10
Value 7/10

How Ministral 3 8B compares

Ministral 3 8B ranks #239 of 378 AI models we track for overall intelligence, #239 of 315 for coding, #214 of 289 for agentic tasks. At $0.15 per million input tokens it is cheaper than 57% of comparable models.

Performance Indices

Source: Artificial Analysis

14.8 Intelligence Index
10 Coding Index
15.6 Agentic Index
31.7 Math Index

Benchmark Scores

Intelligence

GPQA Diamond 47.1% Graduate-level scientific reasoning
HLE 4.3% Humanity's Last Exam
MMLU Pro 64.2% Multi-task language understanding
AIME 2025 31.7% Competition mathematics (2025)
SciCode 20.8% Scientific computing

Technical

LiveCodeBench 30.3% Live coding evaluation
TerminalBench Hard 4.5% Agentic terminal tasks
τ²-Bench 26.6% Conversational agent benchmark

Content

IFBench 29.1% Instruction following
LCR 24% Long-context reasoning

Benchmark data from Artificial Analysis and Hugging Face

How does Ministral 3 8B stack up?

Compare side-by-side with other professional models.

Compare Models

Model Information

ProviderMistral
Release Date December 2, 2025
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.15 $0.000150
Output $0.15 $0.000150

Leaderboard Categories

Frequently asked questions about Ministral 3 8B

How much does Ministral 3 8B cost?

Ministral 3 8B costs $0.15 per million input tokens and $0.15 per million output tokens.

Is Ministral 3 8B good for coding?

On our coding benchmark index, Ministral 3 8B ranks #239 of 315 models, placing it in the broader range of the field for code generation and debugging.

Who created Ministral 3 8B?

Ministral 3 8B is developed by Mistral and was released on December 2, 2025.