Qwen2.5-Math-72B: benchmark results
Provider: Alibaba. Access: Open.
Unified ELO 1474 ± 1, rank #817 of 1392 rated models, from 25 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MERA - SimpleAr | 99.6 | EM (%) | 80.3 |
| MERA - ruModAr | 83.17 | EM (%) | 77.5 |
| MERA - ruMultiAr | 44.14 | EM (%) | 76.6 |
| MERA - MathLogicQA | 73.23 | Accuracy (%) | 75.6 |
| MERA - BPS | 98.4 | Accuracy (%) | 75.5 |
| PhysicsFinals | 32.2 | Score (self-reported) | 75 |
| MERA - PARus | 88.4 | Accuracy (%) | 68.9 |
| MERA - ruMMLU | 67.34 | Accuracy (%) | 62.7 |
| MERA - ruHateSpeech | 77.74 | Accuracy (%) | 58.4 |
| MERA - MaMuRAMu | 74.58 | Accuracy (%) | 56.7 |
| MERA - ruOpenBookQA | 86.25 | Accuracy (%) | 55.5 |
| MERA | 0.51 | Total score | 54.8 |
Interactive version: theaggregate.ai/model?slug=qwen2-5-math-72b · How It Works · Data refreshed daily, snapshot 2026-09-05.