Open LMM Reasoning - MathVista - ARI — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-78B | 74.5 |
| 2 | InternVL3-38B | 73.7 |
| 3 | Doubao-1.5-Pro | 72 |
| 4 | InternVL3-14B | 68 |
| 5 | Grok 2 (1212) | 67.1 |
| 6 | Ovis2-8B | 66.6 |
| 7 | InternVL2.5-78B | 66.6 |
| 8 | GPT-4o ChatGPT | 66.6 |
| 9 | GPT-4.1 (2025-04-14) | 65.7 |
| 10 | InternVL3-8B | 64.6 |
| 11 | GPT-4.1 Mini | 64 |
| 12 | QVQ-72B-Preview | 64 |
| 13 | Claude 3.7 Sonnet | 62.9 |
| 14 | Qwen 2 VL 72B | 62.6 |
| 15 | Gemini 2.0 Flash | 61.2 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mathvista-ari · How the rankings work · Data refreshed daily, snapshot 2026-07-22.