Chaos Cypher
Local-first knowledge graph engine with GraphRAG search and a 46-tool MCP server
- Pricing
- Open source
- Platforms
- self-hosted, CLI, Web
- Runs
- Locally
- Version
- 0.7.2
- Listing updated
- 3 days ago
About Chaos Cypher
Chaos Cypher is an open-source MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.
The MCP interface is one part of a larger self-hosted engine. Chaos Cypher extracts entities and relationships from documents, audio, video, images and web pages (30+ formats) into a graph you can explore and correct, then lets assistants query it. The README says the MCP server exposes 46 tools for search, traversal and graph building to Claude Desktop, Cursor, ChatGPT and other clients.
Search fuses graph traversal with vector search using Personalized PageRank and Reciprocal Rank Fusion, with keyword, semantic and hybrid modes as alternatives. Answers cite the sources and connections behind them. Extraction and chat can run fully offline through Ollama or use OpenAI, Anthropic or Gemini, configured per operation.
By default only the LLM calls you configure leave your machine; embeddings are computed locally. Knowledge can be exported as a portable .ccx package. The vendor's own benchmark on 22 local models reports fewer failed answers than plain vector RAG; those are the vendor's figures. It installs from PyPI as chaoscypher-cli or runs as a Docker container.
Install
Run in a terminal
claude mcp add chaoscypher -- uvx chaoscypher-cliAdd to claude_desktop_config.json
{
"mcpServers": {
"chaoscypher": {
"command": "uvx",
"args": [
"chaoscypher-cli"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"chaoscypher": {
"command": "uvx",
"args": [
"chaoscypher-cli"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"chaoscypher": {
"type": "stdio",
"command": "uvx",
"args": [
"chaoscypher-cli"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"chaoscypher": {
"command": "uvx",
"args": [
"chaoscypher-cli"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"chaoscypher": {
"command": "uvx",
"args": [
"chaoscypher-cli"
]
}
}
}What it does
- Inspectable graph
- Explore entities and relationships on a filterable graph canvas.
- GraphRAG search
- Combines graph traversal with vector search plus keyword and hybrid modes.
- Cited chat
- Answers trace back to sources and connections.
- Local or cloud LLMs
- Use Ollama offline or OpenAI, Anthropic and Gemini per operation.
- Portable packages
- Export sources, graph and settings as one .ccx file.
- Quality scoring
- Rates graph richness 0-100 and flags weak sources.
- Automations
- Visual workflow builder with triggers and conditions.
Good for
- Let Claude query a private document collection as a graph
- Trace an answer back to its source passages
- Build a knowledge base from PDFs, audio and video offline
- Share a knowledge graph with a teammate as a package
Pricing
Open source
Open-source and self-hosted
Prices change. Check chaoscypher.com before you buy.
Chaos Cypher FAQ
Does my data leave my machine?
By default only the LLM calls you configure do; embeddings are local, and a local Ollama model keeps everything on-device.
Which formats can it ingest?
More than 30, including PDF, DOCX, Markdown, HTML, EPUB, audio, video, images and ZIP archives.
How many MCP tools are there?
The README says 46 tools for search, traversal and graph building.
How do I run it?
Install `chaoscypher-cli` from PyPI or run the Docker image from ghcr.io.
Is Chaos Cypher free?
Yes. Chaos Cypher is open source, so you can use and self-host it for free.
How do I add Chaos Cypher to Claude Code?
Run this command in your terminal: claude mcp add chaoscypher -- uvx chaoscypher-cli
What are the best alternatives to Chaos Cypher?
The closest alternatives to Chaos Cypher are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.
Alternatives to Chaos Cypher

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Claude Code Ultimate Guide
MCP serverMCP server for searching the Claude Code Ultimate Guide from any MCP client
Runs locally- Open source
- 6.5K GitHub stars
- AI models and LLM tooling

Nautilus Compass Memory
MCP serverLocal-first long-term memory layer for AI agents, served over MCP
Not checked yet- 5 tools
- Remote
- Open source
- 1.3K GitHub stars
- Knowledge and memory

note1
MCP serverMCP server giving AI assistants search and summaries across your note1 meetings
- 4 tools
- Remote
- Free trial
- 1 GitHub star
- Productivity
MedXpert 图书馆连接器(通用版)
MCP serverLocal MCP server that searches and answers questions over your own Markdown wiki
Runs locally- 4 tools
- Open source
- Knowledge and memory