Agent-FriendlyPopularBuilder FavoriteOpen Source

Memory

Persistent knowledge graph memory across AI conversations.

Setup: 2 minLow complexityTop MCPs for AI Agents

What it does

Maintains a persistent knowledge graph of entities and relationships stored locally. The model can create, query, update, and delete memories across sessions.

Why it matters

AI models are stateless by default. Memory MCP is the simplest way to give an agent long-term knowledge that carries over between sessions.

Best for

  • Long-running agent workflows
  • Personal AI assistants
  • Project tracking
  • User preference storage
  • Agent knowledge accumulation

Not ideal for

  • Single-session tasks
  • Structured database queries
  • High-volume data storage

When to use it

When building agents that need to remember user preferences, ongoing projects, past decisions, or any information that should persist over time.

When not to use it

For single-session tasks where persistence is not needed, or when using a database MCP for structured storage is more appropriate.

Key features

  • Persistent knowledge graph
  • Entity and relationship storage
  • Cross-session memory
  • Local file storage
  • Official Anthropic support

Frequently asked questions

Where is memory stored?

In a local JSON file on your machine. It persists between sessions and is only accessible to the MCP.

Can the model forget memories?

Yes. The model can delete specific entities or observations from the knowledge graph.

Install

Install Memory
$ npx -y @modelcontextprotocol/server-memory

Scores

Ease of setup
5/5
Flexibility
4/5
Popularity
5/5
Agent useful
5/5
Builder speed
5/5

Details

Pricing
open source
Setup time
2 min
Complexity
Low

Works with

ClaudeAny MCP-compatible client

Alternatives

More AI Agents MCPs

Other tools in the same category worth evaluating.

Agent-Friendly

Structured step-by-step reasoning for complex problem solving.

reasoningthinkingagentsplanning
2 minLow

Exploring Top MCPs for AI Agents? See all AI Agents MCPs →