Medical MM Leaderboard - SLAKE: leaderboard
The SLAKE slice of the Medical Multimodal Leaderboard — semantically labeled medical visual question answering with external knowledge; accuracy (%).
Metric: Accuracy (%). Source: huggingface.co. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MedGemma-4B-IT | 76.4 | #853 |
| 2 | Gemini 2.5 Flash | 75.8 | #245 |
| 3 | InternVL3-8B | 72.8 | #620 |
| 4 | InternVL3-14B | 72.8 | #505 |
| 5 | GPT-4.1 | 72.2 | #248 |
| 6 | InternVL3-38B | 71 | #405 |
| 7 | Claude Sonnet 4 | 70.6 | #199 |
| 8 | Qwen 2.5 VL 7B | 47.4 | #643 |
Interactive version: theaggregate.ai/benchmark?slug=medical-mm-leaderboard-slake · How It Works · Data refreshed daily, snapshot 2026-10-20.