Notifications
Clear all
LangGraph Reviews
1
Posts
1
Users
0
Reactions
4
Views
Topic starter
21/07/2026 4:08 am
I'm building my first LangGraph agent that needs to process a large PDF document. The idea is to have one node do the initial parsing, then pass the content to a second node for summarization.
But if I pass the full document text in the state dictionary between nodes, I'm worried it'll overload the memory, especially for big files. I'm using the Python library.
What's the standard way to handle this? Should I be storing the content somewhere else (like a vector store) and just passing a reference? Or is there a built-in way to manage large data in the state?