# Fraise

> Fraise is an open-source MCP server for knowledge and memory. An in-memory temporal graph for agent memory, queried with two verbs: remember and recall. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** macOS, Linux, CLI, API, self-hosted
- **Website:** https://docs.getfraise.dev
- **Repository:** https://github.com/fraisehq/fraise
- **MCP Registry name:** io.github.FraiseHQ/fraise
- **Categories:** knowledge and memory, databases
- **Listing updated:** 2026-11-14

## About

Fraise is a memory database that agents query directly in FQL, a small language with two verbs. A line such as remember 'acme moved to annual billing' topic:billing entity:acme stores a fact, and recall billing entity:acme since:30d top:5 returns the best few matches. Facts link to the entities they mention and the topics they are about, forming a temporal graph.

Recall ranks across full-text, graph and optional vector indexes in a single query and favours recent facts by default. It ships as a single binary installed with Homebrew, Linux packages, Docker or go install, and it exposes an HTTP API at port 9876. The MCP entry runs from a container image and talks to a running Fraise daemon.

The README is candid about maturity: persistence is the next major piece of work, with forget planned in a later release, and it calls the project good for building agent memory today but not yet for long-term production. Its LoCoMo comparison against other memory systems is the vendor's own benchmark run.

## Features

- **Two verbs:** remember and recall cover storing and retrieving.
- **Hybrid retrieval:** Text, graph and optional vector search ranked together.
- **Recency ranking:** Recent facts outrank older ones by default.
- **Single binary:** No separate database to provision.
- **HTTP API:** Query over a local HTTP endpoint.
- **Open source:** MIT licensed.

## Good for

- Give an agent memory across sessions
- Recall what changed for a customer in the last 30 days
- Run low-latency memory lookups mid-conversation
- Compare memory systems on LoCoMo with the open harness

## Install

### Claude Code (Terminal)

```sh
claude mcp add fraise -- docker run -i --rm ghcr.io/fraisehq/fraise:0.3.7
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "fraise": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/fraisehq/fraise:0.3.7"
      ]
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "fraise": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/fraisehq/fraise:0.3.7"
      ]
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "fraise": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/fraisehq/fraise:0.3.7"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "fraise": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/fraisehq/fraise:0.3.7"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "fraise": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/fraisehq/fraise:0.3.7"
      ]
    }
  }
}
```

## FAQ

### How do I store and fetch a memory?

Use remember with a statement plus topic and entity tags, then recall with filters like since and top.

### Is data persisted?

Not yet. The README says persistence is the next major piece of work.

### What licence is it under?

MIT.

### Is it ready for production?

The README says it is good for building agent memory today but not yet for long-term production use.

### Is Fraise free?

Yes. Fraise is open source, so you can use and self-host it for free.

### How do I add Fraise to Claude Code?

Run this command in your terminal: claude mcp add fraise -- docker run -i --rm ghcr.io/fraisehq/fraise:0.3.7

### What are the best alternatives to Fraise?

The closest alternatives to Fraise are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and Midplane.

## Alternatives

- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Nautilus Compass Memory](https://ai-powered.com/mcp/nautilus-compass-memory): Nautilus Compass Memory is an open-source MCP server for knowledge and memory.
- [Midplane](https://ai-powered.com/mcp/midplane): Midplane is an open-source MCP server for databases.
- [note1](https://ai-powered.com/mcp/note1): note1 is an MCP server for productivity.
- [MedXpert 图书馆连接器（通用版）](https://ai-powered.com/mcp/medxpert): MedXpert 图书馆连接器（通用版） is an open-source MCP server for knowledge and memory.
- [Context Nest](https://ai-powered.com/mcp/context-nest): Context Nest is a freemium MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/fraise
