MMLU-by-task - College Chemistry: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 1257 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Mistral-7B-v0.1 | 52 |
| 2 | Llama 2 70B Base | 51 |
| 3 | LLaMA-65B | 49 |
| 4 | Llama 2 70B Chat (HF) | 48 |
| 5 | Llama 2 70B Chat | 46 |
| 6 | vicuna-13B | 44 |
| 7 | LLaMA-30B | 43 |
| 8 | vicuna-13B-v1.5 | 41 |
| 9 | ChatGLM2 6B | 40 |
| 10 | CodeLlama-13B-hf | 40 |
| 11 | GPT-NeoXT-Chat-Base-20B | 39 |
| 12 | CodeLlama-34B-hf | 39 |
| 13 | CodeLlama-7B-Python-hf | 39 |
| 14 | Llama 2 13B Chat Base | 38 |
| 15 | vicuna-13B-v1.3 | 38 |
Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-college-chemistry · How It Works · Data refreshed daily, snapshot 2026-09-05.