Hella swag — Epoch AI
Hella swag — Epoch AI is run by Epoch AI. It has ranked 113 models, of which the catalogue holds 35, scoring from 0.4 to 0.892 on Overall accuracy.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 3rdof 113 | Llama 3.1 405B | Overall accuracy | 0.892 |
| 4thof 113 | falcon-180B | Overall accuracy | 0.89 |
| 5thof 113 | DeepSeek-V3 | Overall accuracy | 0.889 |
| 6thof 113 | DeepSeek-V2 | Overall accuracy | 0.871 |
| 11thof 113 | falcon-40b | Overall accuracy | 0.853 |
| 13thof 113 | Llama-2-70b-hf | Overall accuracy | 0.853 |
| 17thof 113 | StableBeluga2 | Overall accuracy | 0.841 |
| 22ndof 113 | Qwen 2.5 Coder 32B | Overall accuracy | 0.83 |
| 23rdof 113 | falcon-11B | Overall accuracy | 0.829 |
| 31stof 113 | Phi-3-medium-128k-instruct | Overall accuracy | 0.824 |
| 32ndof 113 | gemma-7b | Overall accuracy | 0.822 |
| 39thof 113 | Mistral-7B-v0.1 | Overall accuracy | 0.81 |
| 40thof 113 | Llama-2-13b | Overall accuracy | 0.807 |
| 49thof 113 | llama-13b | Overall accuracy | 0.792 |
| 54thof 113 | falcon-7b | Overall accuracy | 0.781 |
| 62ndof 113 | Llama-2-7b | Overall accuracy | 0.772 |
| 63rdof 113 | Phi-3-small-8k-instruct | Overall accuracy | 0.77 |
| 65thof 113 | Qwen2.5-Coder-7B | Overall accuracy | 0.768 |
| 66thof 113 | Phi-3-mini-4k-instruct | Overall accuracy | 0.767 |
| 70thof 113 | Yi-9B | Overall accuracy | 0.764 |
| 73rdof 113 | llama-7b | Overall accuracy | 0.762 |
| 76thof 113 | opt-66b | Overall accuracy | 0.745 |
| 77thof 113 | Bloom | Overall accuracy | 0.744 |
| 78thof 113 | Yi-6B | Overall accuracy | 0.744 |
| 79thof 113 | xgen-7b-8k-base | Overall accuracy | 0.742 |
| 82ndof 113 | gemma-2b | Overall accuracy | 0.714 |
| 88thof 113 | gpt-neox-20b | Overall accuracy | 0.705 |
| 90thof 113 | opt-1.3b | Overall accuracy | 0.699 |
| 94thof 113 | gpt-j-6b | Overall accuracy | 0.662 |
| 105thof 113 | phi-2 | Overall accuracy | 0.536 |
| 106thof 113 | Qwen2.5-Coder-0.5B | Overall accuracy | 0.484 |
| 107thof 113 | phi-1_5 | Overall accuracy | 0.476 |
| 110thof 113 | gpt-neo-2.7B | Overall accuracy | 0.427 |
| 112thof 113 | stablelm-tuned-alpha-7b | Overall accuracy | 0.407 |
| 113thof 113 | gpt2-xl | Overall accuracy | 0.4 |
Read from the board on 2026-09-16