Skip to content

MemGPT

Let a fixed-context LLM manage its own memory

Visit official site

What MemGPT is

The system is described as a fixed-context LLM processor enhanced with a hierarchical memory system and a set of functions. It lets the model manage its own memory by moving data between contexts.

What you can do with it

  • Parse LLM text output each cycle
  • Execute function calls from that output
  • Move data between main and external context

Memory notes

  • Main context is a fixed-length input
  • Function calls can return control
  • A yield waits for an external event

Pros & cons

✓ What we like

  • A function call can request immediate chaining
  • External events such as user messages trigger the next run

! What to watch out for

  • Main context has a fixed length
  • The LLM pauses until an external event triggers a run

FAQ

What is the main context?

The card says it is the fixed-length LLM input.

What happens on a yield?

The LLM does not run again until an external event triggers it.

Last reviewed: 2026-11-15

More AI chatbot tools

View all →

How we review