On June 20, 2024, the Peking University team released ChatLaw, which has become a new heavyweight tool for digitalization and AI-powered office work in the legal industry.ChatLaw is designed specifically for the Chinese legal environment, combining the advantages of open source, knowledge graphs, and multi-agent collaboration to promote the intelligentization of office processes such as legal case analysis, document writing, and knowledge retrieval. Completely free and open source, it is suitable for lawyers, corporate legal departments, developers, and general users, helping the legal industry reduce costs, increase efficiency, and achieve digital transformation.
June 20, 2024 – ChatLaw, a large-scale legal model developed by a team led by Peking University, has officially become a key component of the legal industry's digital upgrade and...AI office toolsHeavyweight newcomer to the trackWith the large-scale application of artificial intelligence across various industries, the legal field is also ushering in a revolution driven by big data models. ChatLaw, known for its open-source nature, professionalism, and practicality, is dedicated to providing efficient, intelligent, and reliable legal services for legal professionals, corporate legal departments, and the general public. This report will provide a detailed analysis of its advantages, functions, application methods, and applicable users, and will use checklists and tables to give readers a comprehensive understanding of this new-era AI office tool.
The ChatLaw open-source platform address is:https://github.com/PKU-YuanGroup/ChatLawThe research team, leveraging its strong legal expert resources, rich legal text corpora, and diverse knowledge graph technology, has developed ChatLaw into a forward-looking intelligent office assistant that closely reflects the Chinese legal practice environment.
Main functions of ChatLaw

ChatLaw is designed as a large-scale Chinese legal model based on multi-agent collaboration, focusing on handling Chinese legal terminology and the local regulatory environment. Its core functions include:
1. Mixture-of-Experts (MoE) mechanism
- Dedicated Chinese legal language processingChatLaw uses a MoE structure, where multiple "expert models" make decisions in layers based on the question type to achieve detailed solutions.
- Legal knowledge graph enhancementBy integrating knowledge graphs, ChatLaw is able to identify semantic connections between cases and regulations and match them more accurately.
- Standardized Operating Procedures (SOPs)Simulates law firm collaboration, standardizes processes such as analysis, reasoning, and document generation, significantly reduces the "illusion" of legal AI, and improves reliability.
2. Advanced Text Understanding and Structured Extraction
- Automatic file/recording reading recognitionIt can directly read Chinese legal documents and audio, and automatically extract key information.
- Fact-based analysis and visualizationGenerate case mind maps, charts, and visual reports with one click to improve the efficiency of case sorting and analysis.
- Automatically summarize key pointsFor complex cases and lengthy text materials, it intelligently summarizes legal facts and logical connections.
3. Intelligent legal document writing
- Automatic document writingBased on user needs and case facts, and with fine-tuning based on millions of real documents, it enables one-click generation of documents such as contracts, complaints, answers, and legal opinions.
- Professional and reliable legal advice outputCombining knowledge base and SOPs, provide actionable recommendations that comply with current laws and regulations.
4. Intelligent search for laws and regulations
- High-accuracy text similarity matchingThrough in-depth training on past court judgments, we can accurately match user questions with legal provisions.
- Supports personalized regulatory push notifications: To provide targeted interpretation and explanation of the specific legal provisions involved in the case.

Comparison table of main technologies and functions
| Functional categories | ChatLaw application | Traditional legal search / AI Q&A | Innovation Advantages |
|---|---|---|---|
| Legal text interpretation | MoE Legal Big Data Model and Knowledge Graph | General AI Model | Focusing on Chinese law, embedding prior knowledge |
| Structured fact extraction | One-click mind mapping, automatic fact sorting | Hand sorting | Automated, visual analysis, and high consistency |
| Document Writing and Advice | Million-level corpus training, high-dimensional professional document generation | Templates or human writing | Intelligent summarization, highly professional |
| Legal provisions retrieval and matching | Based on the text similarity correspondence of judicial documents | Simple keyword search | Excellent semantic reasoning ability |
| Error rate / hallucination rate | SOP processes are strictly controlled. | Relying on the general-purpose capabilities of AI | The problem of legal AI illusion has been significantly reduced. |
For a detailed understanding of ChatLaw's architecture and innovative principles, please click [here].Read the official ChatLaw paper。
ChatLaw Pricing & Solutions
Open source solutions and licensing information
ChatLaw, an open-source AI office tool, is currently completely free for individual and organizational users and is licensed under the AGPL-3.0 open-source license.Users can download the source code, model weights, and related toolkits for free on the GitHub platform, and perform secondary development or deployment in their internal environment in accordance with the terms of the agreement.

Fees and Authorization List
| Scheme type | Scope of application | cost | Scale adaptation | Remark |
|---|---|---|---|---|
| Open source individuals | Personal/Research | free | 1-10 people | Fully open source |
| Enterprise self-deployment | Corporate/Law Firm | free | 10+ people | Compliant with AGPL-3.0 license |
| Commercial customization | Third-party developers | Contact the team for a quote | No limit | The agreement must be followed and the source of the software must be clearly indicated. |
Currently, the ChatLaw team is also advancing the commercial ecosystem for enterprise-level scenarios, which you can follow.ChatLaw Cloud Official WebsiteLearn about the latest deployments and payment plans. For example, enterprise-grade SaaS services and API access will be launched in the future.
How to use ChatLaw
ChatLaw's open-source nature means that users can easily get started, customize, and integrate it into their own workflows.The following are the main use cases and steps for ChatLaw:
1. Obtaining and Installing
- Go toGitHub project homepage Clone or download the source code.
- Configure the Python environment, dependencies, and model weights according to the documentation.
- It supports local server deployment and cloud platform deployment (such as Tencent Cloud, Alibaba Cloud, etc.).
2. Startup and Input

- Once the ChatLaw service is launched, users can enter legal questions, materials, documents, or audio via the web, API, or command line.
- Fill in the facts of the case or upload materials, and the system will automatically analyze and extract key information.
- Users can obtain answers to legal questions, draft documents, or case logic diagrams.
3. Output and Secondary Development
- The output can be in various formats, including detailed responses, process-oriented suggestions, and editable documents.
- It can be used for secondary development in conjunction with the ChatLaw API, or integrated into the enterprise's internal OA system to achieve automated and intelligent office work.
ChatLaw Integration and Use Case Summary List
- Law Firm/Legal Department Smart Assistant
- Corporate daily legal management
- Legal technology platform development and secondary integration
- Personal academic/practical research
For detailed operating procedures and technical support, please see [link/reference].ChatLaw User Guide。
Who is ChatLaw for?
ChatLaw is geared towards legal and related professionals across multiple levels and fields, and is particularly suitable for the following user groups:
1. Legal professionals
- Lawyers and legal professionals can quickly organize case materials and assist in the writing of legal documents.
- Educators and law school faculty and students conduct automated case analysis and legal teaching.
2. Corporate and Institutional Legal Services
- Corporate compliance and daily legal management, standard contract drafting, and one-click risk alerts.
- Large corporations are promoting the digitalization of legal affairs and the implementation of ISO process standards.
3. Startups and Technology Developers

- LegalTech product developers can build intelligent legal assistants at low cost.
- Digital government and internet platforms are rapidly integrating AI-powered legal consultation capabilities.
4. Individual users & the general public
- Legal consultations for the public, document drafting, risk identification, and legal education.
- Simulated question answering and case practice for academic research or bar exam preparation users.
ChatLaw Target Audience Table
| User Categories | Main requirements | ChatLaw Advantages |
|---|---|---|
| Lawyer/Legal Affairs | Case analysis and document writing | Automation and high consistency |
| Enterprises/Institutions | Legal management and contract review | High efficiency and compliance |
| Technology developers | AI tool integration and platform development | The model is flexible and open. |
| Individual/Student | Legal consultation, learning, and document generation | One-stop, free |
ChatLaw Technical Architecture and Open Community
ChatLaw employs a high-level large language model architecture, fully integrating knowledge graphs and multi-agent collaboration mechanisms.Currently, the mainstream ChatLaw model versions include:
- ChatLaw2-MoE (Main Version)Based on the InternLM architecture, with a 4x7B expert group, specializing in complex reasoning in legal texts.
- ChatLaw-13B and 33B seriesSuitable for different computing power and scenarios, covering highly complex judicial tasks and rapid question-and-answer functions.
Technical architecture highlights:
- Fine-tuning of domain-specific corpora effectively avoids the "language illusion" of general large models.
- Multi-module collaboration, hierarchical and segmented problem-solving, with the final answer "determined" by the most suitable expert in the field.
- It can continuously expand plugins and external APIs to support flexible use in multiple scenarios.
Frequently Asked Questions
1. How does ChatLaw differ from traditional AI office tools?
ChatLaw uses a specially designedCustomized large-scale models and knowledge graphs in the Chinese legal fieldCompared to general AI office tools (such as general GPT-like assistants), it is better able to understand professional legal terminology and Chinese legal contexts, and its performance and accuracy are far superior, greatly alleviating the industry pain point of the illusion of legal AI.
2. Will ChatLaw leak my case privacy?
ChatLaw itself isOpen source tools for local deploymentData is entirely managed by the user. Deploying the service locally or in a private cloud environment maximizes the privacy and security of cases and materials. For cloud services, it is recommended to choose encrypted transmission and pay attention to the platform's data compliance policy.
3. Is ChatLaw applicable to the domestic judicial examination or academic teaching?
Very suitableChatLaw contains a large number of past bar exam questions, case studies, and legal knowledge points, which can help candidates practice questions and review cases. It is also used by law professors in universities for course interaction and case analysis, thus improving the level of intelligence in legal education.
Driven by breakthroughs in legal technology and innovative applications of AIChatLaw is becoming an important driving force for the digital and intelligent transformation of China's legal system.With the continuous growth of the open-source community and its deeper integration with enterprise applications, ChatLaw is expected to drive the development of intelligent legal office tools in China to a new level. For more information on ChatLaw, including downloads and usage, please visit [website address].GitHub HomepageThis marks the beginning of a new journey in intelligent legal office work.
data statistics
Data evaluation
This site's AI-powered navigation is provided by Miao.ChatLawAll external links originate from the internet, and their accuracy and completeness are not guaranteed. Furthermore, AI Miao Navigation does not have actual control over the content of these external links. As of 5:21 PM on September 25, 2025, the content on this webpage was compliant and legal. If any content on the webpage becomes illegal in the future, you can directly contact the website administrator for deletion. AI Miao Navigation assumes no responsibility.

