MMODELYST
Models/Devstral Small 2
MDL

Devstral Small 2

Mistral

Verified against Artificial Analysis · Jul 13, 2026
Capability
47
percentile index
Price /M
$0 / $0
in / out
Context
Avg score
31.8
10 benchmarks
AA Intelligence
17.4
index
Speed
71 tok/s
TTFT 0.6s
Released
Dec 9, 2025
About

Devstral Small 2 is a proprietary model in the Devstral family from Mistral. Benchmarked on 10 evals, averaging 31.8.

Benchmark scores · 10
avg 31.8 — higher is better
MMLU-Pro
67.8
GPQA Diamond
53.2
LiveCodeBench
34.8
AIME
34.3
IFBench
31.2
SciCode
28.8
AA Long-Context Reasoning
24
τ²-bench
23.4
Terminal-Bench Hard
16.7
Humanity's Last Exam
3.4
The research behind this model
all papers mentioning it →

Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.

Details
PublisherMistral
FamilyDevstral
Parameters
License
WeightsClosed
ReleasedDec 9, 2025
Performance & price
Output speed71 tok/s
Latency (TTFT)0.6s
Price in / out$0 / $0 per 1M
Coding index29.3
Math index34.3

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers
Artificial Analysis$0/M