Open LMM Reasoning - MathVista - STA: leaderboard
The "MathVista - STA" subtask of Open LMM Reasoning. A task from the Open LMM Reasoning leaderboard, which evaluates large multimodal models on multi-step visual, mathematical and logical reasoning problems.
Metric: Accuracy (%). Source: huggingface.co. 100 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | SenseNova | 88.4 | |
| 2 | InternVL3-78B | 87 | #352 |
| 3 | QVQ-72B-Preview | 86.7 | #440 |
| 4 | GPT-4.1-mini-20250414 | 86.4 | |
| 5 | Taichu-VLR-7B | 86.4 | #526 |
| 6 | Qwen2.5-VL-72B | 86 | |
| 7 | GLM-4v-Plus-20250111 | 85.7 | |
| 8 | MUG-U-7B | 85 | |
| 9 | InternVL3-38B | 85 | #405 |
| 10 | InternVL2.5-78B-MPO | 84.7 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-mathvista-sta · How It Works · Data refreshed daily, snapshot 2026-10-20.