The LangMem SDK is a lightweight Python library that helps your agents learn and improve through long-term memory. It provides tooling to extract information from conversations, optimize agent behavior through prompt updates, and maintain long-term memory about behaviors, facts, and events.

How it fits into LangChain's ecosystem:

• LangGraph Checkpointers allow for durable execution & message history

• LangGraph BaseStore provide long-term memory with vector search

• LangMem SDK provides the actual logic for creating, updating, and retrieving memories

Why LangMem is valuable:

• Flexible & modular → Works standalone or with LangGraph, compatible with any storage backend

• Structured updates → Uses trustcall for type-safe memory consolidation & invalidation

• Beyond raw chat history → Captures user preferences, procedures, and policies, making agents more adaptive over time

• Efficient memory management → Avoids reliance on long-context models or purely retrieval-based approaches

We’re also chatting with folks on the waitlist for a hosted/managed memory service. Sign up for the waitlist here.

Read more: https://blog.langchain.dev/langmem-sdk-launch/