Memory

MemGPT

Self-memory architecture that gives LLMs persistent storage and automatic context window management.

Overview

MemGPT enables large language models to manage their own memory by abstracting the context window and automatically paging in/out relevant data. It uses a mix of archival, recall, and core memory to simulate human-like recall over extended interactions.

Best for

experiments with autonomous agents that need long-running state

Trade-offs

vs LangGraph

MemGPT handles memory abstraction natively inside the LLM loop; LangGraph gives you explicit control over state transitions and checkpoints.

What XeroHack pre-wires

When you pick MemGPT in the interview, the engine emits these files into your scaffold:

CLI agent template
OpenAI API wrapper
SQLite/Postgres memory backends

Coupon

No active coupon for this tool right now. Sign in to be notified when one is available.