Open LMM Reasoning - WeMath - Cutting and Combining of Figures — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 83 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Doubao-1.5-Pro | 87.9 |
| 2 | Claude 3.7 Sonnet | 84.6 |
| 3 | GPT-4.1 (2025-04-14) | 82.8 |
| 4 | Gemini 2.0 Flash | 82.2 |
| 5 | InternVL3-14B | 79.3 |
| 6 | InternVL3-38B | 79 |
| 7 | InternVL3-78B | 75.8 |
| 8 | GPT-4o ChatGPT | 75.3 |
| 9 | InternVL3-8B | 75.2 |
| 10 | Gemini 1.5 Pro (002) | 75 |
| 11 | Claude 3.5 Sonnet (20241022) | 73.7 |
| 12 | Qwen 2 VL 72B | 73.3 |
| 13 | QVQ-72B-Preview | 71.4 |
| 14 | InternVL2.5-78B | 70.3 |
| 15 | InternVL2-8B | 69.8 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-wemath-cutting-and-combining-of-figures · How the rankings work · Data refreshed daily, snapshot 2026-07-22.