ZeroEval MATH-500: leaderboard
Allen AI ZeroEval evaluation of MATH-500: competition-level math problems, scored independently from Artificial Analysis.
Metric: MATH-500 Score. Source: huggingface.co. Source: www.vals.ai. Saturation forecast: Estimated already saturated. 33 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | LongCat-Flash-Thinking | 99.2 | #363 |
| 2 | Sarvam 105B | 98.6 | #409 |
| 3 | GLM-4.5 | 98.2 | #275 |
| 4 | GLM 4.5 Air | 98.1 | #392 |
| 5 | NVIDIA Nemotron Nano 9B V2 | 97.8 | #823 |
| 6 | Kimi K2 | 97.4 | #243 |
| 7 | Kimi K2 0905 | 97.4 | #288 |
| 8 | Llama 3.1 Nemotron Ultra 253B v1 | 97 | #441 |
| 9 | Sarvam 30B | 97 | #719 |
| 10 | Qwen 3 14B | 96.8 | #537 |
| 11 | MiniMax M1 80k | 96.8 | #432 |
| 12 | LongCat Flash Lite | 96.8 | #681 |
| 13 | LongCat-Flash-Chat | 96.4 | #331 |
| 14 | Claude 3.7 Sonnet | 96.2 | #249 |
| 15 | MiniMax M1 40k | 96 | #683 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=zeroeval-math-500 · How It Works · Data refreshed daily, snapshot 2026-10-20.