Open LMM Reasoning - MathVista - FQA — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 79.2 |
| 2 | InternVL3-78B | 78.1 |
| 3 | QVQ-72B-Preview | 77.7 |
| 4 | Qwen 2 VL 72B | 76.6 |
| 5 | InternVL3-38B | 75.5 |
| 6 | Doubao-1.5-Pro | 75.5 |
| 7 | GPT-4o ChatGPT | 74.3 |
| 8 | InternVL3-14B | 74 |
| 9 | GPT-4.1 (2025-04-14) | 73.6 |
| 10 | Claude 3.5 Sonnet (20241022) | 72.5 |
| 11 | Ovis2-8B | 70.3 |
| 12 | InternVL2.5-78B | 70.3 |
| 13 | Gemini 2.0 Flash | 69.5 |
| 14 | InternVL3-8B | 69.1 |
| 15 | Qwen 2 VL 7B | 68.8 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mathvista-fqa · How the rankings work · Data refreshed daily, snapshot 2026-07-22.