L

Lingshu 7B

Developed by lingshu-medical-mllm
Lingshu is a SOTA multimodal large language model in the medical field, performing excellently in medical visual Q&A and report generation tasks.
Downloads 355
Release Time : 6/5/2025

Model Overview

The Lingshu model is a general - purpose foundation model focusing on multimodal understanding and reasoning in the medical field, supporting more than 12 medical imaging modalities.

Model Features

Multimodal medical understanding
Supports more than 12 medical imaging modalities, including X - rays, CT scans, MRI, etc.
Excellent performance
Achieves SOTA level in most medical multimodal/text Q&A and report generation tasks
Multi - task support
Supports multiple medical tasks such as visual Q&A, text Q&A, and report generation simultaneously

Model Capabilities

Medical image analysis
Medical text understanding
Medical report generation
Multimodal reasoning
Medical Q&A

Use Cases

Medical diagnosis assistance
Image report generation
Automatically generate diagnostic reports based on medical images
ROUGE - L reaches 30.8 and CIDEr reaches 109.4 on the MIMIC - CXR dataset
Medical visual Q&A
Answer clinical questions based on medical images
Accuracy reaches 67.9% on the VQA - RAD dataset
Medical education
Medical knowledge Q&A
Answer various medical knowledge questions
Accuracy reaches 74.5% on the MMLU - Med dataset
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase