Qwen3 30B A3B Thinking 2507 — Benchmarks

Benchmark scores for Qwen3 30B A3B Thinking 2507 aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #26 of 78 open modelscomposite 52.2/100 across 8 benchmarks in 4 categories · methodology

Coding

BenchmarkScoreOpen rankAll models
SciCode33.3%#45 / 62#140 / 161

Knowledge

BenchmarkScoreOpen rankAll models
MMLU-Pro80.9%#22 / 144#60 / 257

Math

BenchmarkScoreOpen rankAll models
AIME 2024/202570.3%#26 / 82#129 / 270

Reasoning

BenchmarkScoreOpen rankAll models
Chess Puzzles8.0%#26 / 65#138 / 205
DTBench69.3%#32 / 67#135 / 210
LMCA19.5%#33 / 53#137 / 172
GPQA Diamond70.1%#31 / 94#163 / 289
CritPt0.3%#39 / 62#131 / 172

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