Qwen2.5 Coder 0.5B — Benchmarks

Benchmark scores for Qwen2.5 Coder 0.5B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #111 of 123 open modelscomposite 24.5/100 across 5 benchmarks in 3 categories · methodology

Knowledge

BenchmarkScoreOpen rankAll models
HellaSwag48.4%#39 / 42#69 / 76
ARC Challenge34.4%#45 / 51#70 / 77
MMLU42.0%#65 / 77#118 / 136

Math

BenchmarkScoreOpen rankAll models
GSM8K34.5%#39 / 60#57 / 93

Reasoning

BenchmarkScoreOpen rankAll models
WinoGrande54.8%#43 / 46#77 / 80

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