FrontierAI / ML

haiku-rag

io.github.ggozad/haiku-rag · v0.39.0

Opinionated agentic RAG powered by LanceDB, Pydantic AI, and Docling

Repository

Gecko Rating

Frontier

GitHub Stars

658

as of 2026-11-14

Transport

Stdio

v0.39.0

Last Update

2026-04-09

Install haiku-rag

Ready-to-paste config for every major MCP client

claude mcp add haiku-rag -- npx io.github.ggozad/haiku-rag # or uvx io.github.ggozad/haiku-rag for Python

Stdio transport · runs the server as a local subprocess.Claude Code docs

Why Frontier?

500+ GitHub stars · battle-tested, widely adopted, safe default.

Has repo

Has website

Classified

Transport specified

Alternatives

Same category · same transport · ranked by stars

More AI / ML servers

Top picks in the same category

mcp-server-browserbase

Frontier · 79 gecko

tooluniverse

Frontier · 77 gecko

Embeddable badge

Drop into your README · auto-updates when the rating changes

[![haiku-rag · Gecko Rating](https://benchgecko.ai/badge/mcp/haiku-rag.svg)](https://benchgecko.ai/mcp/haiku-rag)

The badge auto-updates · your tier and star count refresh every 24 h.

FAQ

Auto-generated from the dataset · updated daily

What is haiku-rag?

haiku-rag is an MCP (Model Context Protocol) server in the AI / ML category. Opinionated agentic RAG powered by LanceDB, Pydantic AI, and Docling

How is haiku-rag rated?

BenchGecko rates haiku-rag Frontier on the Gecko Rating ladder (64/100). 500+ GitHub stars · battle-tested, widely adopted, safe default. The underlying GitHub repository has 658 stars.

How do I install haiku-rag?

haiku-rag supports Claude Code, Cursor, Windsurf, and VS Code. Each client needs a different config format · BenchGecko shows ready-to-paste snippets for all four above. Stdio servers run locally via npx; streamable-http and SSE servers connect over the network.