详细内容
Top Searches
qwen3
Z-Image
qwen
deepseek
Qwen3-VL
Z-Image-Turbo
IndexTTS2
qwen2.5
z image
ComfyUi
login / register
Fast-forwarding AI development with open-source community.
Join Us
Models
Tasks
Z-Image-Turbo
2025.11.28
FLUX.2-dev
2025.11.27
DeepSeek-Math-V2
2025.11.27
HunyuanOCR
2025.11.25
Kimi-K2-Thinking
2025.11.08
HunyuanVideo-1.5
2025.11.29
MiMo-Embodied-7B
2025.11.29
DeepSeek-V3.1-Nex-N1
2025.11.19
Computer Vision
Visual inspection and tracking
Optical Character Recognition
Face and body
Visual Classification
Visual Editing
Visual Segmentation
Natural Language Processing
Text Categorization
Text Generation
Participle
Named Entity Recognition
Translation
Text Summarization
Voice
Speech Recognition
Inverse Text Regularization
Speech Synthesis
Voice wake-up
Speech noise reduction
Echo Cancellation
Multimodal
Image Description
Video Description
Visual Positioning
Text to image
Text generation video
Multimodal representation
Scientific Computing
Protein structure generation
Protein function prediction
More →
Datasets
let_dataset
2025.11.30
veAgentBench
2025.11.25
Honey-Data-15M
2025.11.26
pico-banana-400k
2025.10.26
fineweb-edu-100b-shuffle
2025.10.14
More→
Studios
PaddleOCR-VL_Online_Demo
2025.10.16
Wan2.5-Preview-Demo
2025.10.24
More →
Get to know us
Community trends
Online Experience
Open Source
Partners
login /register
Get to know us
Quick Start
Basic concepts and environment setups, bring the model potentials to your fingertips in a minute.
ModelScope
Explore the industry's latest and finest models, datasets, and AI applications. Get access to computation infrastructure for model serving and start building your own models and applications with our open source community.
ModelScope Library
The Python Library that offers efficient model inference, fine-tune, and evaluation. Your unified gateay to unleash potentials of various AI models ranging from vision, speech to LLM.
ModelHub
The open-source hub for hosting AI models, datasets, and more.
This Week's Trending
Models
Z-Image-Turbo\
\
2025.11.287.9k433 Kimi-K2-Thinking\
\
2025.11.08170.7k70 FLUX.2-dev\
\
2025.11.272.6k30 DeepSeek-Math-V2\
\
2025.11.2785023 z_image_turbo\
\
2025.11.2626621 HunyuanOCR\
\
2025.11.258.3k33
Studios
Wan2.2-Animate\
\
2025.11.12217.1k441 Z-Image-Gallery\
\
2025.11.2615.3k37 PaddleOCR-VL_Online_Demo\
\
2025.11.1024.6k56 index-tts-vllm\
\
2025.11.2556.3k160 Wan2.5-Preview-Demo\
\
2025.09.2641.1k80 FinResearch\
\
2025.11.252.6k10
Datasets
let_dataset\
\
2025.11.30160.6k1 financial_classification\
\
2025.11.216.8k4 veAgentBench\
\
2025.11.25373 Honey-Data-15M\
\
2025.11.262.6k12 pico-banana-400k\
\
2025.10.2743417 sam-3d-body-dataset\
\
2025.11.20243
Popular Studio
Tongyi Laboratory
666666
Studios
ModelScope's Studio provides you with a free and flexible AI application display space. You can build and demonstrate different AI applications by yourself based on the atomic capabilities provided by the models on the ModelScope platform.
2025.11.30 updated
Tongyi Laboratory
666666
Studios
ModelScope's Studio provides you with a free and flexible AI application display space. You can build and demonstrate different AI applications by yourself based on the atomic capabilities provided by the models on the ModelScope platform.
2025.11.30 updated
Tongyi Laboratory
666666
Studios
ModelScope's Studio provides you with a free and flexible AI application display space. You can build and demonstrate different AI applications by yourself based on the atomic capabilities provided by the models on the ModelScope platform.
2025.11.30 updated
Tongyi Laboratory
666666
Studios
ModelScope's Studio provides you with a free and flexible AI application display space. You can build and demonstrate different AI applications by yourself based on the atomic capabilities provided by the models on the ModelScope platform.
2025.11.30 updated
PaddlePaddle\
\
24.6k56\
\
PaddleOCR-VL_Online_Demo\
\
2025.11.10 updated 
Wan-AI\
\
41.1k80\
\
Wan2.5-Preview-Demo\
\
2025.09.26 updated
ms-agent\
\
2.6k10\
\
FinResearch\
\
FinResearch：面向金融研究的多智能体分析引擎，是MS-Agent项目下专为金融领域打造的智能研究工作流，旨在自动化实现从原始市场信号到专业级研究报告的端到端生成。它不再依赖单一大模型“幻觉式”输出，而是通过结构化任务分解 + 多源异构数据融合 + 可验证分析闭环，实现高可信度、高解释性、高可复现性的金融洞察.\
\
2025.11.25 updated 
Qwen\
\
31.7k114\
\
Qwen3-VL-Demo\
\
2025.09.24 updated
xmccln\
\
17.7k96\
\
DeepSeek-OCR-demo\
\
DeepSeek-OCR 是一种革命性的OCR技术，它将OCR从一个简单的“文字提取”工具，升级为一个高效的“文档理解”系统。它通过“光学压缩”技术，将图像直接转换为LLM可高效处理的紧凑视觉Token，极大地提升了处理长文档的速度和效率，为大规模AI训练数据的生成、知识库构建和聊天机器人上下文管理等领域带来了新的可能性\
\
2025.11.14 updated 
opencompass\
\
306.9k261\
\
CompassArena\
\
CompassArena 司南大模型竞技场收集了近期最流行的大语言模型和多模态模型，用户可以轻松体验和比较多款主流大模型的效果。新一代CompassArena新增 Judge Copilot 功能，榜单算法全新升级，同时新增 30+ 全新模型，欢迎体验！\
\
2025.11.03 updated
OpenDataLab\
\
187.6k256\
\
MinerU\
\
一站式开源高质量数据提取工具\
\
2025.11.04 updated
Comprehensive open-sourced framework
to assist model development
and building of AI applications
AllModelScopeEval-ScopeSwiftModelScope-Agent
ModelScope
让“模型即服务”的理念生动实现！
Bring the notion of Model-as-a-Service to life.
Eval-Scope
一个用于简化大模型评估和性能基准测试的高效、可定制框架。
A streamlined and customizable framework for efficient large model evaluation and performance benchmarking.
Swift
魔搭大模型训练推理工具箱，支持LLaMA、千问、ChatGLM、BaiChuan等多种模型及LoRA等多种训练方式。
The LLM training/inference framework of ModelScope community, Support various models like LLaMA, Qwen, ChatGLM, Baichuan and others, and training methods like LoRA, ResTuning, NEFTune, etc.
ModelScope-Agent
连接 ModelScope 模型能力与万物的桥梁。
An agent framework connecting models in ModelScope with the world.
Partners
Model Partners
Framework Partners
Join our partnership, contact us for cooperation.
About ModelScope was founded by Institute for Intelligent Computing in June 2022, in conjunction with CCF Open Source Development Committee, as a open source model community and innovation platform. By partnering with prestigious institutes and industry leaders, ModelScope is committed to building an open community that inspires innovation and promotes development of AI applications. It is tasked with the mission to bring the concept "Model-as-a-Service" to life, and to build a prosperous AI ecosystem around open-source models.
Code of ConductTerms-and-ConditionPrivacy-PolicyProposal of Artificial Intelligence Ethics
Contact usE-mail
DingTalk
© 2022-2025 ModelScope.cn all rights reserved