# DeepMem

Persistent memory for AI agents: facts are written once and recalled by meaning. Useful when context keeps vanishing between sessions.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 114
- Author: deepmemteam
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/deepmemteam/deepmem
- Updated: 2026-08-17
- HTML: https://affpapa.org/en/nevibecode/mcp/deepmem

## What it does

DeepMem is an HTTP memory server that speaks the Mem0 API. Under the hood: FastAPI in front of a Qdrant vector store, LLM fact extraction, hybrid retrieval (vector plus BM25, entity boost, time-decay), semantic caching and async batched distillation. Multi-tenant isolation comes from the `user_id` in the request body.

Two ways in. The client library: `pip install deepmem-client`, then swap one import line — `from mem0 import MemoryClient` becomes `from deepmem import MemoryClient`. Or run it as an MCP server: in Claude Code that is `claude mcp add --transport http deepmem https://deepmem.dev/mcp`. A `dm_live-...` key is only needed for the cloud; the self-hosted server runs in open mode with no key and no registration.

Some behaviour is worth knowing up front. With `add(infer=True)` the write is asynchronous — you get `pending=True` and an empty results list, and extracted facts land a few seconds later. There are no graph relations, so `relations` is always empty; retrieval is hybrid, not graph-based. `reset` is scoped to a single `user_id` and asks for confirmation.

## Requirements
- Remote server: `https://deepmem.dev/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http deepmem https://deepmem.dev/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add deepmem --url https://deepmem.dev/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "deepmem": {
            "url": "https://deepmem.dev/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "deepmem": {
            "type": "http",
            "url": "https://deepmem.dev/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http deepmem https://deepmem.dev/mcp
```

## When to use it
- recall who the user is and what they were building last session
- keep root causes and dead ends across sessions
- store preferences without restating them every time
- run agent memory on your own box instead of someone else's cloud
- move a project off Mem0 by changing one import line

## Example prompts
- Remember that I'm working on a FastAPI service and prefer pytest
- What do you know about me from previous sessions?
- Search your memory for the error we debugged last time

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 386989
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33918
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81115
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119160
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47392
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43872

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md