What is this?
LangChain is a framework for building agents and LLM-powered applications: it chains together interoperable components and third-party integrations to simplify AI app development.
LangChain is a framework for building agents and LLM-powered applications: it chains together interoperable components and third-party integrations to simplify AI app development.Do I code it into my own app, or is it a hosted platform?You code against it: an open-source Python framework you import into your project, with an equivalent JS/TS library, LangChain.js, for the JavaScript side.
You code against it: an open-source Python framework you import into your project, with an equivalent JS/TS library, LangChain.js, for the JavaScript side.How do I go from zero to a first working agent?The Quickstart routes you by ambition: start with Deep Agents for common agent patterns, or move to LangGraph when you need advanced customization and orchestration.
The Quickstart routes you by ambition: start with Deep Agents for common agent patterns, or move to LangGraph when you need advanced customization and orchestration.Can I use it with models like OpenAI, Anthropic, or Gemini?Yes — integrations are the core idea: chat and embedding models plus tools and toolkits, and the repo lists OpenAI, Anthropic, and Gemini among its providers.
Yes — integrations are the core idea: chat and embedding models plus tools and toolkits, and the repo lists OpenAI, Anthropic, and Gemini among its providers.Is everything I might need inside this one package?No — the framework works standalone, but it is one piece of a larger ecosystem: Deep Agents, LangGraph, and LangSmith are separate packages and products you adopt as needed.
No — the framework works standalone, but it is one piece of a larger ecosystem: Deep Agents, LangGraph, and LangSmith are separate packages and products you adopt as needed.Is this for quick experiments or serious production apps?Both: it is built for rapid prototyping with modular components, and it grows into production with battle-tested patterns and built-in monitoring and evaluation support.
Both: it is built for rapid prototyping with modular components, and it grows into production with battle-tested patterns and built-in monitoring and evaluation support.What is Deep Agents, the suggested starting point?A higher-level package built on LangChain: its agents ship with built-in planning, subagents, and file system usage, covering common agent patterns out of the box.
A higher-level package built on LangChain: its agents ship with built-in planning, subagents, and file system usage, covering common agent patterns out of the box.When would I reach for LangGraph instead?LangGraph is the low-level side of the ecosystem: a framework for building controllable agent workflows that reliably handle complex tasks, below Deep Agents' higher-level abstractions.
LangGraph is the low-level side of the ecosystem: a framework for building controllable agent workflows that reliably handle complex tasks, below Deep Agents' higher-level abstractions.What is LangSmith and will I need it?It is the companion product for developing, debugging, and deploying agents — evals, observability, and debugging for LLM apps, plus a deployment platform for long-running, stateful workflows.
It is the companion product for developing, debugging, and deploying agents — evals, observability, and debugging for LLM apps, plus a deployment platform for long-running, stateful workflows.What are similar projects?langgraph, PocketFlow, agentic-rag-for-dummies, docker-agent, langflow, mem0, langchain4j, voltagent
langgraph, PocketFlow, agentic-rag-for-dummies, docker-agent, langflow, mem0, langchain4j, voltagentWhich topics is it in?deepagents, llm-framework, ai-framework, multiagent, pydantic, langgraph, langchain, embeddings, enterprise, chatgpt, orchestration, agent-orchestration
deepagents, llm-framework, ai-framework, multiagent, pydantic, langgraph, langchain, embeddings, enterprise, chatgpt, orchestration, agent-orchestration