What Is Mems

Mems is an industrial-grade memory hub solution for AI agents. Through a four-layer hot/cold-decoupled architecture, it provides a memory foundation with personality consistency, low-cost retrieval, and structured evolution.

It addresses a few recurring problems:

prompt windows are short and recent context disappears quickly
raw chat logs pile up without becoming stable long-term knowledge
current session state, historical events, preferences, and archive data are often mixed together
third-party systems struggle to integrate with unclear memory APIs

Mems solves this with four explicit layers:

L0: live session context
L1: episodic memory and online history
L2: distilled long-term knowledge
L3: durable JSONL archive

For integrators, the main flow is simple:

GET /v1/mems/context
POST /v1/mems/query
let the agent build the answer
POST /v1/mems/write