Easy interaction with LLMs and Vector Stores
All major commercial and open-source LLMs and Vector Stores are easily accessible through a unified API, enabling you to build chatbots, assistants and more.
Tailored for Java
Smooth integration into your Java applications is made possible thanks to Quarkus, Spring Boot and Helidon integrations. There is two-way integration between LLMs and Java: you can call LLMs from Java and allow LLMs to call your Java code in return.
Agents, Tools, RAG
Our extensive toolbox provides a wide range of tools for common LLM operations, from low-level prompt templating, chat memory management, and output parsing, to high-level patterns like Agents and RAG.
LangChain4j - 为 Java 开发者打造的大语言模型（LLM）应用开发框架
LangChain4j 是面向 Java 生态的大语言模型应用开发框架，提供统一 API 接入 OpenAI、Azure OpenAI、Google Gemini、Anthropic Claude、通义千问、DeepSeek、文心一言、智谱 GLM、Ollama 等 20+ 主流大模型，以及 Milvus、Pinecone、Weaviate、Qdrant、Elasticsearch、PgVector 等 30+ 向量数据库。
内置 RAG 检索 增强生成、Agents 智能体、Tools 工具调用（Function Calling）、MCP（Model Context Protocol）、AI Services、Chat Memory、Guardrails 等高层能力，并与 Spring Boot、Quarkus、Helidon、Micronaut 深度集成，帮助 Java 工程师快速构建可上生产的 AI 应用。