Open LMM Reasoning - MathVista - VQA — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o ChatGPT | 64.2 |
| 2 | InternVL3-78B | 63.7 |
| 3 | InternVL3-38B | 62.6 |
| 4 | Doubao-1.5-Pro | 61.5 |
| 5 | Qwen 2 VL 72B | 59.8 |
| 6 | Gemini 2.0 Flash | 59.2 |
| 7 | GPT-4.1 (2025-04-14) | 58.1 |
| 8 | Qwen 2 VL 7B | 57 |
| 9 | InternVL2.5-78B | 57 |
| 10 | GPT-4.1 Mini | 55.3 |
| 11 | Ovis2-8B | 55.3 |
| 12 | InternVL3-14B | 54.7 |
| 13 | QVQ-72B-Preview | 53.6 |
| 14 | InternVL3-8B | 53.1 |
| 15 | Gemini 1.5 Pro (002) | 52.5 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mathvista-vqa · How the rankings work · Data refreshed daily, snapshot 2026-07-22.