Falcon 180B — Benchmarks
Benchmark scores for Falcon 180B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Overall rank: #16 of 123 open modelscomposite 66.6/100 across 8 benchmarks in 3 categories · methodology
Knowledge
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| HellaSwag | 89.0% | #2 / 42 | #4 / 76 |
| ARC Challenge | 67.8% | #19 / 51 | #27 / 77 |
| BoolQ | 89.0% | #2 / 33 | #6 / 77 |
| OpenBookQA | 64.2% | #8 / 21 | #12 / 42 |
| MMLU | 70.6% | #25 / 77 | #59 / 136 |
Math
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| GSM8K | 54.4% | #28 / 60 | #41 / 93 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| WinoGrande | 87.1% | #2 / 46 | #5 / 80 |
| PIQA | 84.9% | #3 / 35 | #5 / 60 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.