Mistral: Mistral Small 3

Mistral: Mistral Small 3

mistralai · Released Jan 30, 2025 Efficient
Intelligence #227 / 599
40.0 Our Score
Speed #61 / 281
156.7 tokens / sec
Input #159 / 599
$0.050 per 1M tokens
Output #140 / 599
$0.080 per 1M tokens
Context #414 / 599
32,768 tokens

Analysis Summary

Mistral Small 3 is a compact model from Mistral AI priced at $0.05 input and $0.08 output per million tokens, making it one of the most affordable options available. Its agentic index of 19.6 is disproportionately high relative to its intelligence index of 6.9, suggesting reasonable tool-orchestration capability for its size. MMLU-Pro of 0.652 and LiveCodeBench of 0.252 are modest.

For businesses, Mistral Small 3 suits high-volume, cost-sensitive workflows such as content classification, simple drafting, or lightweight tool-calling pipelines. Its math index of 4.3 and limited coding benchmarks make it unsuitable for technical or analytical tasks. The 32K context window is a constraint for long-document work.

At this price point, it is a strong fit for teams running large-scale automation where per-token cost is the dominant concern. Pair it with a more capable model for tasks requiring deeper reasoning.

Assessed June 30, 2026

Editorial notes

Mistral Small 3 is an cost-efficient model with a surprisingly strong agentic index, suited to high-volume lightweight tasks where budget is the primary constraint.

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

Performance Profile

Intelligence1.9Technical2.5Value7.8Content4.5
Intelligence 1.9/10
Technical 2.5/10
Content 4.5/10
Value 7.8/10

How Mistral: Mistral Small 3 compares

Mistral: Mistral Small 3 ranks #288 of 387 AI models we track for overall intelligence, #185 of 294 for agentic tasks. Its 33K-token context window is larger than 31% of the models we list. At $0.05 per million input tokens it is cheaper than 73% of comparable models.

About Mistral: Mistral Small 3

Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed..

24B Parameters

Performance Indices

Source: Artificial Analysis

6.9 Intelligence Index
19.6 Agentic Index
4.3 Math Index

Benchmark Scores

Intelligence

GPQA Diamond 46.2% Graduate-level scientific reasoning
HLE 4.1% Humanity's Last Exam
MMLU Pro 65.2% Multi-task language understanding
MATH 500 71.5% Mathematical problem-solving
AIME 8% Competition mathematics
AIME 2025 4.3% Competition mathematics (2025)
SciCode 23.6% Scientific computing

Technical

LiveCodeBench 25.2% Live coding evaluation
τ²-Bench 19.6% Conversational agent benchmark

Content

IFBench 26.4% Instruction following

Benchmark data from Artificial Analysis and Hugging Face

How does Mistral: Mistral Small 3 stack up?

Compare side-by-side with other efficient models.

Compare Models

Model Information

OpenRouter ID mistralai/mistral-small-24b-instruct-2501
Providermistralai
Model FamilyMistral
Release Date January 30, 2025
Context Length32,768 tokens
Max Completion16,384 tokens
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.05 $0.000050
Output $0.08 $0.000080

Live Performance

Live endpoint metrics, refreshed every 30 minutes.

100%
Avg Uptime
257ms
Best Latency (TTFT)
47 tok/s
Best Throughput
1/1
Active Endpoints
Available via: DeepInfra

Leaderboard Categories

Frequently asked questions about Mistral: Mistral Small 3

How much does Mistral: Mistral Small 3 cost?

Mistral: Mistral Small 3 costs $0.05 per million input tokens and $0.08 per million output tokens.

What is the context window of Mistral: Mistral Small 3?

Mistral: Mistral Small 3 has a context window of 32,768 tokens (33K).

Who created Mistral: Mistral Small 3?

Mistral: Mistral Small 3 is developed by Mistral and was released on January 30, 2025.