LangChain

Overview

LangChain is a framework for building agents and LLM-powered applications. It provides create_agent, a minimal, highly configurable agent harness. The harness is everything around the model loop: the prompt, the tools, and any middleware that shapes behavior. LangChain supports OpenAI, Anthropic, Google, and more. Its agents are built on top of LangGraph, and LangSmith can trace, debug, and evaluate agent behavior.

Key Features

• Standard model interface for chat models, embeddings, and more across providers, letting teams switch models with minimal code changes.
• create_agent provides a minimal, highly configurable harness composed from model, tools, prompt, and middleware.
• Built on top of LangGraph, enabling durable execution, human-in-the-loop support, persistence, and more.
• LangSmith integration for tracing requests, debugging agent behavior, and evaluating outputs.
• Real-time data augmentation by connecting LLMs to diverse data sources and external or internal systems.
• Model interoperability and rapid prototyping through a modular, component-based architecture.

Use Cases

• Building agents and LLM-powered applications.
• Composing a customizable agent harness from model, tools, prompt, and middleware.
• Connecting LLMs to diverse data sources and external or internal systems for real-time data augmentation.
• Swapping models across providers to find the best choice for application needs.

Getting Started

• Install LangChain with uv add langchain.
• Initialize a chat model with init_chat_model and invoke it, for example with openai:gpt-5.5.
• See the Installation instructions and Quickstart guide to start building agents and applications.
• Use LangSmith to trace requests, debug agent behavior, and evaluate outputs by setting LANGSMITH_TRACING=true and your API key.

Before You Adopt

• License: MIT. Review its terms before using, modifying, or distributing the project.
• Tracing, debugging, and evaluation capabilities rely on LangSmith and an API key, an external dependency for observability.
• LangChain agents depend on LangGraph as the underlying orchestration layer for durable execution, human-in-the-loop support, and persistence.