Building LangChain and LangGraph 1.0

LangChain CEO Harrison Chase sits down with open source engineers Sydney, Hunter, and Will for an in-depth technical discussion on the major 1.0 releases of LangChain and LangGraph. The team explores the evolution to production-ready agent frameworks, including the new create_agent abstraction, middleware system, and why controllability matters for building reliable AI applications.

Learn how to build with LangChain and LangGraph on LangChain Academy: https://bit.ly/3Lhaj3q
Check out the docs: https://bit.ly/42MylJi
Learn more about LangChain & LangGraph 1.0 in the blog post: https://bit.ly/4qkNhs7

00:00 - Introductions and team overview
01:15 - The origins of LangChain
03:00 - Why we built LangGraph
05:30 - What's in an agent runtime
08:45 - Rewriting LangChain on top of LangGraph for 1.0
10:20 - The create_agent abstraction in LangChain 1.0
12:00 - Middleware in LangChain 1.0
14:30 - Pre-built middleware for summarization, human-in-the-loop
16:15 - The end of the "champion model" era
18:00 - Content blocks for standardizing model outputs
20:30 - When to use LangChain vs LangGraph (agents vs workflows)
23:00 - Context engineering and composability with middleware
25:15 - What's next for LangChain
27:30 - Closing thoughts and getting started with 1.0