Qwen3.5 27B — Benchmarks

Benchmark scores for Qwen3.5 27B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #10 of 78 open modelscomposite 58.8/100 across 4 benchmarks in 3 categories · methodology

Coding

BenchmarkScoreOpen rankAll models
WeirdML39.5%#23 / 46#108 / 161
LMArena WebDev1357.8#33 / 48#92 / 123

Instruction Following

BenchmarkScoreOpen rankAll models
LMArena Text1407.9#41 / 182#133 / 391

Knowledge

BenchmarkScoreOpen rankAll models
MMLU-Pro86.1%#5 / 144#28 / 257

Reasoning

BenchmarkScoreOpen rankAll models
DTBench82.4%#20 / 67#101 / 210
LMCA34.0%#16 / 53#106 / 172

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