OpenVLM MathVista - LOG: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 284 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 56.8 |
| 2 | GPT-4.1 (2025-04-14) | 54.1 |
| 3 | GPT-5 Mini (2025-08-07) | 54.1 |
| 4 | GPT-5 | 51.4 |
| 5 | GPT-4.1 Mini | 51.4 |
| 6 | GPT-5 Nano | 48.6 |
| 7 | Claude 3.5 Sonnet (20241022) | 45.9 |
| 8 | GPT-4o ChatGPT | 43.2 |
| 9 | InternVL3-38B | 37.8 |
| 10 | Doubao-1.5-Pro | 37.8 |
| 11 | InternVL3-78B | 35.1 |
| 12 | Claude 3.5 Sonnet | 35.1 |
| 13 | Grok 2 (1212) | 35.1 |
| 14 | QVQ-72B-Preview | 35.1 |
| 15 | VLM-R1-3B-Math-0305 | 35.1 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mathvista-log · How It Works · Data refreshed daily, snapshot 2026-09-05.