haiku.rag-slim is a Python package for local-first agentic retrieval-augmented generation over user documents. It supports document ingestion, embeddings, hybrid search, reranking, citations, and multimodal retrieval without requiring a database server.
In the RAG, search & retrieval space, haiku.rag-slim takes a focused approach. It focuses on retrieving reliable, cited information from personal documents without operating a database server. It is built as an open-source project for developers building local document retrieval and agent applications. The project is open source (MIT). haiku.rag-slim is available on the command line and API, and it can be self-hosted.
It is developed by ggozad, and it first shipped in 2025. Development happens publicly on GitHub with 584 stars and 704 commits in the last 90 days. Among its 8 catalogued features are document ingestion, hybrid search, and reranking. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do