Open LMM Reasoning - MathVerse - Solid Geometry: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Doubao-1.5-Pro | 52.9 |
| 2 | Claude 3.7 Sonnet | 38.7 |
| 3 | InternVL3-78B | 37.8 |
| 4 | InternVL3-38B | 37.8 |
| 5 | Gemini 1.5 Pro (002) | 37.8 |
| 6 | GPT-4.1 Mini | 34.5 |
| 7 | Claude 3.5 Sonnet (20241022) | 34.5 |
| 8 | GPT-4o ChatGPT | 33.6 |
| 9 | GPT-4.1 (2025-04-14) | 30.3 |
| 10 | InternVL3-8B | 28.6 |
| 11 | QVQ-72B-Preview | 28.6 |
| 12 | Gemini 2.0 Flash | 27.7 |
| 13 | VLAA-Thinker-Qwen2.5VL-7B | 27.7 |
| 14 | InternVL2.5-78B-CoT | 23.5 |
| 15 | Grok 2 (1212) | 22.7 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mathverse-solid-geometry · How It Works · Data refreshed daily, snapshot 2026-09-05.