a-mem-mcp vs Memary
Memory and knowledge graphs for AI agents compared
a-mem-mcp and Memary both add persistent memory and connected knowledge to AI agents. a-mem-mcp focuses on coding-project notes through MCP and a Python API, while Memary combines graph retrieval with user-focused memory in an autonomous-agent application.

a-mem-mcp: Evolving Memory for Coding Agents
A-MEM MCP brings the A-MEM self-evolving memory approach to coding agents, organizing project knowledge as a connected graph stored in ChromaDB. It is currently tested with Claude Code and suits agents that need to recall and connect past project decisions.

Memary: Add Memory to Autonomous Agents with LlamaIndex
Memary is an open-source memory layer for autonomous agents, built around LlamaIndex, knowledge graphs, and a ReAct chat agent. It stores interaction-derived entities and user context to help agents retrieve relevant information and tailor responses.
| a-mem-mcp | Memary | |
|---|---|---|
| Language | Python | Jupyter Notebook |
| License | MIT | MIT |
| Stars | 35 | 2.7k |
| Forks | 3 | 205 |
| Last analyzed | Oct 28, 2026 | Oct 25, 2026 |
Key differences
- a-mem-mcp stores project-specific notes and evolves their links and descriptions; Memary tracks entities across interactions, including their frequency and recency.
- a-mem-mcp provides an MCP server and Python API; Memary includes a ReAct-based chat agent and tools for extending it.
- a-mem-mcp uses ChromaDB for persistence; Memary lists FalkorDB and Neo4j, as well as local JSON files.
- a-mem-mcp is written in Python and reports 35 stars; Memary is listed as Jupyter Notebook and reports 2.7k stars.
- a-mem-mcp identifies Claude Code as its currently tested MCP-compatible agent; Memary describes its included ReAct agent as a demo-oriented starting point.
- Both projects use the MIT license, and both support Ollama; a-mem-mcp also lists SGLang and OpenRouter, while Memary lists graph-database options and external-search fallback.
Choose a-mem-mcp if you…
- need persistent, connected notes about coding projects across sessions.
- want MCP tools or a Python API for adding and retrieving agent memory.
- prefer ChromaDB persistence with configurable LLM backends, including local Ollama.
Choose Memary if you…
- want an agent to personalize responses using entities and prior interactions.
- need graph-based, recursive or multi-hop retrieval, with external-search fallback.
- want to prototype a ReAct agent with user-separated contexts and customizable tools.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.