# Levirge Brain

> Levirge Brain is a paid MCP server for knowledge and memory. Shared MCP memory for AI agents, with a cited knowledge base, skills and handoffs. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Paid
- **Starting price:** $20 a month
- **Platforms:** API, Web
- **Website:** https://levirge.com/brain
- **Repository:** https://github.com/levirge/brain
- **MCP Registry name:** com.levirge/brain
- **Categories:** knowledge and memory, AI models and LLM tooling
- **Listing updated:** 2026-11-14

## About

Levirge Brain aims at the gap between one agent's memory and a team's. Notes are held outside any single tool, model or teammate, so Claude Code, Codex and other MCP clients all read one endpoint. Each project has a Vault shared by several people and their agents, every document records whose agent captured it, and tokens can be limited to chosen Vaults or made read-only.

The README lists search-first memory using semantic search with reranking, write-back with `add_knowledge`, fact lineage back to source documents, session commands such as `/brain:start` and `/brain:finish`, and a feedback signal that tunes retrieval. Agents can hand each other work through a durable inbox with context attached. Statements in summaries link to their sources and summaries rebuild when a source changes. An Obsidian plugin syncs a local vault both ways without overwriting your notes on conflict.

A privacy choice sits at the centre of the model. The Private lane runs all model calls on Levirge-hosted LLMs, while Fast uses third-party cloud models. Plans start at $20 a month with unlimited agents and vaults and are billed by people, storage and lane. Enterprise self-hosting is a roadmap item.

## Features

- **Shared Vaults:** One knowledge base per project, read and written by teammates and their agents.
- **Agent handoffs:** A durable inbox passes work and context between sessions and agents.
- **Cited summaries:** Statements link to sources and rebuild when sources change.
- **Fact lineage:** Trace an answer back to the documents it came from.
- **Obsidian sync:** Two-way sync with a local vault that never overwrites your note on conflict.
- **Private or fast lane:** Choose Levirge-hosted models or third-party cloud models.
- **Scoped tokens:** Limit tokens to chosen Vaults and make them read-only.

## Good for

- Give Codex the decisions you made in Claude Code
- Let a colleague's agent reuse your finished investigation
- Hand failing benchmark defects from one agent to another
- Capture runbooks and gotchas once for every future session
- Curate team knowledge in Obsidian while agents recall it

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http brain https://brain.levirge.com/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://brain.levirge.com/mcp
```

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

```json
{
  "mcpServers": {
    "brain": {
      "url": "https://brain.levirge.com/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "brain": {
      "type": "http",
      "url": "https://brain.levirge.com/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "brain": {
      "serverUrl": "https://brain.levirge.com/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "brain": {
      "type": "http",
      "url": "https://brain.levirge.com/mcp"
    }
  }
}
```

## Tools

- `search_knowledge`: Searches the knowledge base with semantic search and reranking.
- `add_knowledge`: Captures decisions and fixes into the knowledge base.
- `feedback`: Records helpful or dead-end verdicts that tune retrieval.

## FAQ

### What does it cost?

The site says plans start at $20 a month, and $60 a month puts your corpus on the private model lane.

### Which clients work?

Claude Code, Codex, Cline and any MCP client that supports a remote server with OAuth sign-in.

### Where does my content go?

On the Private lane, model calls run on Levirge-hosted LLMs with no third-party provider. The Fast lane sends content to a cloud API.

### Can I self-host?

Self-hosting is described as an enterprise roadmap item.

### Is Levirge Brain free?

No. Levirge Brain is a paid product. Paid plans start at $20 a month.

### How do I add Levirge Brain to Claude Code?

Run this command in your terminal: claude mcp add --transport http brain https://brain.levirge.com/mcp

### What are the best alternatives to Levirge Brain?

The closest alternatives to Levirge Brain are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.

## 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.
- [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.
- [Tulsa Ecosystem Map](https://ai-powered.com/mcp/tulsa-ecosystem-map): Tulsa Ecosystem Map is a free MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/levirge-brain
