Lingshu: MLLMs for Unified Multimodal Medical Understanding and Reasoning

Welcome to the Lingshu project - A Generalist Foundation Model for Unified Multimodal Medical Understanding and Reasoning.

Highlights:

• BIG NEWS: Lingshu Training Data is released for public use.
• BIG NEWS: Lingshu-I-8B is released with state-of-the-art performance on medical VQA tasks. It undergoes the same training pipeline on top of InternVL3.
• Lingshu supports more than 12 medical imaging modalities, including X-Ray, CT Scan, MRI, Microscopy, Ultrasound, Histopathology, Dermoscopy, Fundus, OCT, Digital Photography, Endoscopy, and PET.
• Lingshu models achieve SOTA on most medical multimodal/textual QA and report generation tasks for 7B and 32 model sizes.
• Lingshu-32B outperforms GPT-4.1 and Claude Sonnet 4 in most multimodal QA and report generation tasks.

Quick links:

• Models are available in multiple model sizes: Lingshu-7B, Lingshu-32B
• Paper: Lingshu: A Generalist Foundation Model for Unified Multimodal Medical Understanding and Reasoning
• Paper: ReasonMed: A 370K Multi-Agent Generated Dataset for Advancing Medical Reasoning
• MedEvalKit is available here: MedEvalKit, we are continuously working on it to involve more benchmarks and models.
• Data: ReasonMed
• Project Page: https://alibaba-damo-academy.github.io/lingshu/
• Lingshu FastMCP Medical AI Service is available here: Lingshu_MCP