Open LMM Reasoning - LogicVista — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 83 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 64.9 |
| 2 | GPT-4o ChatGPT | 64.4 |
| 3 | Doubao-1.5-Pro | 64.2 |
| 4 | GPT-4.1 (2025-04-14) | 61.1 |
| 5 | Claude 3.5 Sonnet (20241022) | 60.4 |
| 6 | Claude 3.7 Sonnet | 58.2 |
| 7 | QVQ-72B-Preview | 58.2 |
| 8 | InternVL3-38B | 57.7 |
| 9 | InternVL3-78B | 55.9 |
| 10 | Gemini 1.5 Pro (002) | 54.4 |
| 11 | Gemini 2.0 Flash | 52.3 |
| 12 | InternVL3-14B | 52.1 |
| 13 | Grok 2 (1212) | 49.9 |
| 14 | InternVL2.5-78B | 49 |
| 15 | Gemma 3 27B | 47.9 |
Interactive version: theaggregate.ai/benchmark?slug=open-lmm-reasoning-logicvista · How the rankings work · Data refreshed daily, snapshot 2026-07-22.