Tongyi Farui product overview
Tongyi Farui is a Large Language Model (LLM) designed specifically for the legal industry. Built on Qwen, it is trained on extensive legal domain data and knowledge. By integrating model fine-tuning, reinforcement learning, retrieval-augmented generation (RAG), legal Agents, and judicial domain-specific small models, Tongyi Farui supports features such as answering legal questions, reasoning about legal applicability, recommending similar court cases, assisting with case analysis, generating legal documents, retrieving legal knowledge, and reviewing contract clauses.
|
Service name |
Scenarios |
Description |
|
Tongyi Farui |
Legal consultation |
Provides answers and advice on legal questions, including interpreting specific legal issues, explaining statutes, offering potential solutions, suggesting legal actions, and guiding users through relevant legal procedures. Enhanced training covers causes of action such as marriage and family disputes, private lending, and more. |
|
Case retrieval |
Performs similar-case searches in a database of court judgments based on user input. Automatically analyzes search results and provides relevance assessments of similar cases based on the disputed issues and core facts of the target case. |
|
|
Statute and regulation retrieval |
Searches a legal database based on user input using semantic understanding of legal provisions. Supports both precise and fuzzy searches. |
|
|
Text reading |
Performs legal-level reading comprehension of provided text, identifying legal entities and their relationships, along with timelines of legal events. |
|
|
Document generation |
Generates legal documents automatically by combining user-provided legal requests with an understanding of templates and key elements. Input can be either a case description or a prior legal document. |
|
|
Contract review |
Reviews contracts for clause accuracy, legal compliance, business logic, and potential threats based on predefined rules. Supports custom review rules. |