URIAL-Bench - Math — leaderboard
Metric: Judge Score (0-10). Source: huggingface.co. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4 | 6.8 |
| 2 | GPT-3.5 Turbo | 6.3 |
| 3 | DBRX | 5.9 |
| 4 | Mixtral 8x7B (v0.1) | 4.85 |
| 5 | Yi 34B (Base) | 4.8 |
| 6 | gemma-7B | 4.35 |
| 7 | Phi-2 | 3.8 |
| 8 | Llama 2 70B Base | 3.6 |
| 9 | Mistral-7B-v0.1 | 3.4 |
| 10 | gemma-2B | 3.3 |
| 11 | Llama 2 13B Base | 2.85 |
| 12 | Yi 6B (Base) | 2.5 |
| 13 | Llama 2 7B Base | 1.6 |
| 14 | falcon-7B | 1.6 |
| 15 | mpt-7B | 1 |
Interactive version: theaggregate.ai/benchmark?slug=urial-bench-math · How the rankings work · Data refreshed daily, snapshot 2026-07-22.