This is an early release preview. You may encounter bugs.
Tool
Zep logo
Unclaimed

memory

Zep

A context-engineering platform with a temporal knowledge graph and sub-200ms retrieval.

C 58/100 GitHub score ? This grade is derived from GitHub signals — not user votes. Open for the full breakdown.
No votes yet

01 / About

About Zep.

Zep is a managed agent-memory platform (Zep Cloud) for giving LLM applications and agents persistent, queryable memory. It is built on a temporal knowledge graph (the open-source Graphiti framework) that tracks how facts and relationships evolve over time, and it advertises low-latency retrieval and MCP support, positioning it as a context-engineering layer that supplies relevant history and knowledge back to an agent at inference time. Official SDKs are provided for Python, TypeScript/JavaScript, and Go.

This particular repository is not the product itself but a companion collection of example apps, framework integration packages, default ontology definitions, plugins, memory benchmarks (such as LoCoMo and LongMemEval), and an evaluation harness for ingestion and retrieval. The integrations cover many agent frameworks, including Google ADK, Microsoft Agent Framework, AutoGen, AG2, CrewAI, LangGraph, LiveKit, Pydantic AI, Mastra, and the Vercel AI SDK, each released independently. The earlier self-hostable Zep Community Edition is deprecated and has been moved to a legacy folder.

02 / Discussion CREDIBILITY-GATED

Discussion · 0

Reading is open to everyone. Only verified humans or builders at GitHub B+ can post or vote — every thread carries its author's credibility.

  • No discussions yet — start the first thread.

03 / Build

Build with Zep.

Browse the catalogue for frameworks, tools, and blueprints — each scored on real GitHub credibility.

Browse the catalogue