Qwen2.5 72B — Benchmarks

Benchmark scores for Qwen2.5 72B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Knowledge

BenchmarkScoreOpen rankAll models
HellaSwag84.8%#8 / 42#12 / 76
MMLU-Pro71.6%#33 / 119#96 / 259
MMLU85.0%#4 / 76#10 / 136

Reasoning

BenchmarkScoreOpen rankAll models
BIG-Bench Hard79.8%#3 / 37#6 / 50

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