Devstral Small 2: benchmark scores, price and ranking

Devstral Small 2 from Mistral AI ranks #592 of 743 LLMs on the LLMs Tiger aggregate benchmark score (45.4), based on 2 benchmark results captured up to 2026-10-13.

Organisation: Mistral AI
Type: Open-weight / local model
Open weights: Yes
Released: 2025-12-09 (source: epoch)
Aggregate rank: #592 of 743 (score 45.4)

Across the 2 benchmarks where at least ten models are scored, Devstral Small 2 is in the top 10% on 0 and in the top half on 0.

Devstral Small 2 benchmark results

Devstral Small 2 benchmark scores, ranked against all models measured on each benchmark
Benchmark	Score	Rank	Source	Captured
CritPt	0.0%	#89 of 115	Epoch AI Benchmarking Hub	2026-10-13
SciCode	28.8%	#91 of 106	Epoch AI Benchmarking Hub	2026-10-13

About Devstral Small 2

Devstral Small 2 is an open-weight model that can run on your own hardware from Mistral AI. It was released on 2025-12-09, according to Epoch AI. No API price is listed for it yet. We hold 2 benchmark results for it across 2 areas, and none is in the top tenth of its board.

Where Devstral Small 2 stands

On our aggregate score, Devstral Small 2 is #592 of 743 ranked models, in the lower half. Its score is 45.4; the scale is centred on 50, so it sits 4.6 points below the typical model. That beats 20% of the ranked field. Among the 44 ranked Mistral AI models it is #26. The next one up is mistral-medium-2505 at 45.4. The next one down is Mistral-Nemo-Instruct-2407 at 45.2.

Devstral Small 2, benchmark by benchmark

Devstral Small 2 lands below the middle on CritPt: 0.0%, #89 of 115, with 77% of the field scoring higher. GPT-5.6 Sol (max) leads at 32.3%; the closest result above is DeepSeek-V3 (Mar 2025) at 0.0%. It covers unpublished research-level physics problems. The board median is 2.9%, so Devstral Small 2 is 2.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

SciCode is a weak spot. Devstral Small 2 scores 28.8%, #91 of 106, behind 85% of the models measured. The leader, Claude Opus 5.5 (max), reaches 66.9%. It covers scientific research code. The board median is 44.3%, so Devstral Small 2 is 15.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Strengths and weaknesses

Devstral Small 2 has too few results on boards with ten or more models to say where it is strong or weak. Treat any single score as a data point, not a verdict.

Release history

The Mistral AI model released before it was Mistral Large 3 on 2025-12-02, which scores 42.7. The next one was Mistral Small 4 on 2026-03-16, at 46.5.

Reading these numbers

The 2 scores on this page come from 1 source: Epoch AI Benchmarking Hub (2). No official leaderboard supplied any of them, so treat them as third-party measurements. With 2 results the aggregate rank can move a lot when one new score arrives.

Models ranked nearby

#589 Llama-3-Smaug-8B (45.4)
#590 grok-build-0.1 (45.4)
#591 mistral-medium-2505 (45.4)
#593 Command R+ (45.4)
#594 Qwen2.5-Coder-1.5B-Instruct (45.4)
#595 Llama 3.1-8B (45.3)