Mistral Small 3.1 24B Instruct 2503 — Benchmarks

Benchmark scores for Mistral Small 3.1 24B Instruct 2503 aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #70 of 78 open modelscomposite 22.7/100 across 7 benchmarks in 3 categories · methodology

Coding

BenchmarkScoreOpen rankAll models
SciCode26.5%#52 / 62#149 / 161

Instruction Following

BenchmarkScoreOpen rankAll models
LMArena Text1303.1#101 / 182#261 / 391

Math

BenchmarkScoreOpen rankAll models
MATH Level 546.8%#16 / 39#65 / 107
AIME 2024/20255.8%#65 / 82#238 / 270

Reasoning

BenchmarkScoreOpen rankAll models
Chess Puzzles1.0%#43 / 65#179 / 205
DTBench58.6%#51 / 67#172 / 210
GPQA Diamond47.5%#57 / 94#225 / 289
CritPt0.0%#56 / 62#162 / 172

Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.