Glossary
LangChain
The earliest widely-adopted LLM agent and RAG orchestration framework (2022), now with the LangGraph extension for stateful multi-step agent workflows.
The library that catalyzed the LLM-as-application-substrate ecosystem. LangChain originated in 2022 as Python wrappers over LLM APIs, vector stores, and document loaders; by 2024 it had grown into the dominant agent orchestration framework, with LangGraph layering stateful multi-step workflows on top.
LangChain has been criticized for over-abstraction and a churning API; the company’s response has been LangChain Core (smaller, more stable) plus opinionated subprojects (LangGraph for agents, LangSmith for observability). The framework remains the most-installed agent library and the default lesson example in most LLM courses.
Full coverage at /projects/langchain.