LangChain is an open-source framework in Python and JavaScript for building LLM-based applications such as chatbots, agents, and RAG systems. It provides standardized chains (and LCEL) to compose model calls, tools, memory, and retrievers, as well as ready-made architectures for LLM agents that can autonomously choose and invoke external functions, APIs, and data providers. Thanks to its modular design and broad ecosystem of integrations with model providers, vector databases, and data services, LangChain simplifies rapid prototyping, switching between providers, and moving to production. Companion projects - such as LangGraph (agent orchestration), LangSmith (tracing, observability, evaluations), and LangServe (API deployment)—help build reliable, production-grade AI systems.