Devstral Small 2

Devstral Small 2

Mistral · Released Dec 9, 2025 Emerging
Intelligence #360 / 557
25.1 Our Score
Speed #167 / 257
67.1 tokens / sec
Input
Not priced
Output
Not priced
Context
Not reported

Analysis Summary

Devstral Small 2 sits in the Emerging tier on our leaderboard, ranked #360 of 557 published models on overall intelligence. At $0.000 input and $0.000 output per 1M tokens, it is among the most expensive on the market.

Editorial notes

Devstral Small 2 from Mistral is a coding-focused small model with a 20.7 coding index; useful for lightweight code assistance but outclassed by larger models for complex development tasks.

Assessed May 14, 2026

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

Performance Profile

Intelligence3.7Technical3.2Value0Content3
Intelligence 3.7/10
Technical 3.2/10
Content 3/10
Value 0/10

Performance Indices

Source: Artificial Analysis

19.5 Intelligence Index
20.7 Coding Index
20 Agentic Index
34.3 Math Index

Benchmark Scores

Intelligence

GPQA Diamond 53.2% Graduate-level scientific reasoning
HLE 3.4% Humanity's Last Exam
MMLU Pro 67.8% Multi-task language understanding
AIME 2025 34.3% Competition mathematics (2025)
SciCode 28.8% Scientific computing

Technical

LiveCodeBench 34.8% Live coding evaluation
TerminalBench Hard 16.7% Agentic terminal tasks
τ²-Bench 23.4% Conversational agent benchmark

Content

IFBench 31.2% Instruction following
LCR 24% Long-context reasoning

Benchmark data from Artificial Analysis and Hugging Face

How does Devstral Small 2 stack up?

Compare side-by-side with other emerging models.

Compare Models

Model Information

ProviderMistral
Release Date December 9, 2025
Status Active

Leaderboard Categories