WHAT IT DOES

What does it do?

raggy is a RAG command-line tool for local documents built with LangChain, Chroma, and Ollama. Indexing that combines vector search and BM25, along with embedding generation, runs locally. Answer generation can use either a local LLM or a remote API.

KEY FEATURES

Key features

• Hybrid search combining vector search and BM25 indexing
• Return document-based answers with source locations and relevance scores
• Automatically process images and scanned documents with OCR
• Generate answers with a local LLM or a remote API-based LLM
• Automatically download configured models and reindex documents
• Manage input paths, database location, and runtime options in a configuration file