Meta Llama 3 8B Instruct — Benchmarks
Benchmark scores for Meta Llama 3 8B Instruct aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Overall rank: #103 of 123 open modelscomposite 28.3/100 across 9 benchmarks in 3 categories · methodology
Knowledge
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| MMLU-Pro | 41.0% | #81 / 119 | #197 / 259 |
| ARC Challenge | 82.8% | #10 / 51 | #16 / 77 |
| OpenBookQA | 82.6% | #4 / 21 | #6 / 42 |
| TriviaQA | 67.7% | #13 / 18 | #31 / 39 |
| MMLU | 68.8% | #27 / 77 | #66 / 136 |
Math
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| AIME 2024/2025 | 1.9% | #66 / 74 | #255 / 270 |
| MATH Level 5 | 6.1% | #29 / 32 | #104 / 108 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| GPQA Diamond | 26.1% | #76 / 83 | #280 / 291 |
| WinoGrande | 65.0% | #32 / 46 | #61 / 80 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.