U-MATH - Precalculus — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 28 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O1 | 96.25 |
| 2 | O3 Mini | 94.38 |
| 3 | DeepSeek R1 | 93.12 |
| 4 | QwQ 32B-Preview | 92.5 |
| 5 | O1 Mini | 92.5 |
| 6 | Gemini 2.0 Flash (01-21) (Thinking) | 92.5 |
| 7 | QVQ-72B-Preview | 91.25 |
| 8 | DeepSeek V3 | 86.88 |
| 9 | Gemini 1.5 Pro | 86.25 |
| 10 | Athene-V2-Chat | 85 |
| 11 | Phi-4 | 84.38 |
| 12 | Qwen2.5-Math-72B-Instruct | 84.38 |
| 13 | Qwen 2.5 32B Instruct | 80 |
| 14 | Qwen 2.5 72B Instruct | 79.38 |
| 15 | Gemini 1.5 Flash | 79.38 |
Interactive version: aibenchmarks.dev/benchmark?slug=u-math-precalculus · How the rankings work · Data refreshed daily, snapshot 2026-07-08.