Open LMM Reasoning - MathVision - metric geometry - angle: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Doubao-1.5-Pro | 67.6 |
| 2 | Gemini 1.5 Pro (002) | 52.6 |
| 3 | Gemini 2.0 Flash | 52 |
| 4 | Claude 3.7 Sonnet | 47.4 |
| 5 | Gemma 3 27B | 45.7 |
| 6 | InternVL3-78B | 44.5 |
| 7 | InternVL3-38B | 39.9 |
| 8 | GPT-4.1 (2025-04-14) | 39.9 |
| 9 | InternVL3-8B | 39.3 |
| 10 | GPT-4o ChatGPT | 38.7 |
| 11 | GPT-4.1 Mini | 37.6 |
| 12 | Claude 3.5 Sonnet (20241022) | 36.4 |
| 13 | InternVL2.5-78B | 36.4 |
| 14 | InternVL2.5-8B-BoN-8 | 35.8 |
| 15 | InternVL2.5-78B-CoT | 35.8 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mathvision-metric-geometry-angle · How It Works · Data refreshed daily, snapshot 2026-09-05.