Llama 3.1 405B — Benchmarks

Benchmark scores for Llama 3.1 405B 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
HellaSwag89.2%#1 / 42#3 / 76
MMLU-Pro61.6%#51 / 119#136 / 259
MMLU84.4%#7 / 76#14 / 136

Reasoning

BenchmarkScoreOpen rankAll models
BIG-Bench Hard82.9%#2 / 37#4 / 50

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