Falcon 40B — Benchmarks

Benchmark scores for Falcon 40B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #43 of 73 open modelscomposite 41.2/100 across 4 benchmarks in 3 categories · methodology

Knowledge

BenchmarkScoreOpen rankAll models
HellaSwag85.3%#7 / 42#11 / 76
MMLU56.9%#49 / 76#98 / 136

Math

BenchmarkScoreOpen rankAll models
GSM8K25.0%#47 / 59#68 / 93

Reasoning

BenchmarkScoreOpen rankAll models
BIG-Bench Hard37.1%#29 / 37#41 / 50

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