List your company

Medical Multimodal Geometric Perception on Med-Scout-Bench

94.4Task A ScoreMed-Scout
1938.57558.1577.725Jan 30, 2026

Evaluation results

MethodLinks
Med-ScoutBase Model=QWEN3-VL-4B-INSTRUCT2026.0194.477.589.884.4
Med-ScoutBase Model=QWEN3-VL-8B-INSTRUCT2026.0186.778.190.283.6
Med-ScoutBase Model=LINGSHU-7B2026.0178.977.56071.9
Med-ScoutBase Model=HUATUOGPT-VISION-7B2026.0176.679.36473.8
GEMINI-3-FLASHModel Type=Proprietary2026.016758.876.766.1
GPT-5Model Type=Proprietary2026.0164.356.174.663.6
HUATUOGPT-VISION-7BModel Type=Medical2026.0160.733.72836.3
LINGSHU-7BModel Type=Medical2026.0160.629.121.931.9
QWEN3-VL-4B-INSTRUCTModel Type=General-purpose2026.0159.636.631.338.7
QWEN2.5-VL-3B-INSTRUCTModel Type=General-purpose2026.0146.934.69.228.2
MEDGEMMA-4B-ITModel Type=Medical2026.0146.433.52934.1
QWEN3-VL-8B-INSTRUCTModel Type=General-purpose2026.014134.746.439.7
INTERNVL3-8BModel Type=General-purpose2026.0137.734.327.332.5
QWEN2.5-VL-7B-INSTRUCTModel Type=General-purpose2026.0131.428.127.828.5
LLAVA-MED-7BModel Type=Medical2026.0121.917.48.615.2