OpenVLM MM-Vet - Math: leaderboard
MM-Vet Math score from the OpenCompass OpenVLM Leaderboard (VLMEvalKit): GPT-4-graded open-ended answers on the MM-Vet questions that require arithmetic or equation solving on image content. A question can require several capabilities, so the capability subsets overlap. Score 0-100, higher is better.
Metric: Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 284 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4.1 (2025-04-14) | 95.4 | #257 |
| 2 | GPT-4.5 | 95.4 | #278 |
| 3 | GPT-4o ChatGPT | 95.4 | |
| 4 | InternVL3-78B | 91.9 | #361 |
| 5 | Gemini 2.5 Pro | 91.5 | #159 |
| 6 | InternVL3-14B | 91.5 | #506 |
| 7 | GPT-5 | 88.1 | #105 |
| 8 | InternVL3-38B | 88.1 | #413 |
| 9 | Gemini 2.0 Flash | 87.7 | #351 |
| 10 | InternVL3-8B | 87.3 | #624 |
| 11 | GPT-4.1 Mini | 84.2 | #363 |
| 12 | Gemini 1.5 Pro (002) | 83.8 | #368 |
| 13 | Gemma 3 27B | 83.1 | #606 |
| 14 | Claude 3.5 Sonnet (20241022) | 80.4 | #302 |
| 15 | GPT-5 Mini (2025-08-07) | 80.4 | #175 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mm-vet-math · How It Works · Data refreshed daily, snapshot 2026-11-05.