Medical Visual Understanding on GMAI-MMBench
61.74AccuracyGMAI-VL
Evaluation Results
| Method | Links | |
|---|---|---|
| GMAI-VLParams=7B, Medical=yes2025.10 | 61.74 | |
| UniMedVLParams=14B, Medical=yes2025.10 | 60.75 | |
| Gemini-2.5-FlashParams=Closed, Medical=no2025.10 | 59.05 | |
| GPT-4oParams=Closed, Medical=no2025.10 | 57.27 | |
| Lingshu-32BParams=32B, Medical=yes2025.10 | 52.7 | |
| Lingshu-7BParams=7B, Medical=yes2025.10 | 52.3 | |
| HuatuoGPT-Vision-7BParams=7B, Medical=yes2025.10 | 50.22 | |
| BagelParams=14B, Medical=no2025.10 | 48.11 | |
| Claude Sonnet 4Params=Closed, Medical=no2025.10 | 44.26 | |
| MedGemma-4BParams=4B, Medical=yes2025.10 | 44 | |
| InternVL2Params=8B, Medical=no2025.10 | 43.47 | |
| HealthGPT-L14Params=14B, Medical=yes2025.10 | 43.1 | |
| HealthGPT-M3Params=3.8B, Medical=yes2025.10 | 42.08 | |
| JanusParams=1.3B, Medical=no2025.10 | 39.3 | |
| LLaVA-v1.5Params=7B, Medical=no2025.10 | 38.23 | |
| RadFMParams=14B, Medical=yes2025.10 | 22.34 | |
| LLaVA-MedParams=7B, Medical=yes2025.10 | 20.54 | |
| Med-FlamingoParams=8.3B, Medical=yes2025.10 | 12.74 |