Abnormal Identification on PET-Bench
64.23AI ScoreGemini-2.5-pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini-2.5-proSetting=Zero-shot, Category=Proprietary MLLMs2026.01 | 64.23 | |
| GPT-4oSetting=Zero-shot, Category=Proprietary MLLMs2026.01 | 55.86 | |
| GPT-5Setting=Zero-shot, Category=Proprietary MLLMs2026.01 | 55.49 | |
| MedGemma-4BSetting=Zero-shot, Category=Open-Source Medical MLLMs2026.01 | 53.91 | |
| Qwen3-VL-30BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 53.78 | |
| Shizhen2.5VL-7BSetting=Zero-shot, Category=Open-Source Medical MLLMs2026.01 | 53.14 | |
| Lingshu-32BSetting=Zero-shot, Category=Open-Source Medical MLLMs2026.01 | 52.99 | |
| Lingshu-7BSetting=Zero-shot, Category=Open-Source Medical MLLMs2026.01 | 52.51 | |
| MedGemma-27BSetting=Zero-shot, Category=Open-Source Medical MLLMs2026.01 | 52.14 | |
| InternVL3.5-8BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 51.96 | |
| Shizhen2.5VL-32BSetting=Zero-shot, Category=Open-Source Medical MLLMs2026.01 | 51.3 | |
| InternVL2.5-38BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 51.1 | |
| Qwen2.5-VL-7BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 51.09 | |
| InternVL2.5-8BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 50.9 | |
| Qwen2.5VL-72BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 50.64 | |
| InternVL3-8BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 50.6 | |
| InternVL3-78BSetting=Zero-shot, Category=Open-Source MLLMs2026.01 | 49.17 | |
| Grok-4Setting=Zero-shot, Category=Proprietary MLLMs2026.01 | 49.1 | |
| Claude-sonnet-4.5Setting=Zero-shot, Category=Proprietary MLLMs2026.01 | 48.97 |