OpenVLM MMStar - Logical Reasoning: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 284 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 80.8 |
| 2 | GPT-4o ChatGPT | 80.8 |
| 3 | GPT-5 Mini (2025-08-07) | 80.8 |
| 4 | InternVL3-38B | 79.2 |
| 5 | InternVL2.5-78B | 78.4 |
| 6 | GPT-4.1 (2025-04-14) | 78.4 |
| 7 | Gemini 2.5 Pro | 76.8 |
| 8 | InternVL3-78B | 76.8 |
| 9 | GPT-4.5 | 76.8 |
| 10 | Gemini 2.0 Flash | 75.6 |
| 11 | Claude 3.5 Sonnet (20241022) | 74.4 |
| 12 | GPT-5 Nano | 73.6 |
| 13 | Qwen 2 VL 72B | 72.4 |
| 14 | Grok 2 (1212) | 71.6 |
| 15 | Claude 3.7 Sonnet | 70.8 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmstar-logical-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-05.