rasinmuhammed/rag-tui

Supports multiple chunking strategies (token, sentence, paragraph, recursive, fixed character, custom) with live parameter tuning, and includes a headless CLI and Python API for programmatic access in notebooks and CI pipelines. Integrates with Ollama, OpenAI, Groq, and Google Gemini for semantic retrieval evaluation, and exports production-ready splitter configs directly to LangChain or LlamaIndex formats. Handles diverse file types (.txt, .md, .pdf, .csv, .json, .yaml, .py, .js) and provides batch evaluation metrics like hit rate and relevance scoring to diagnose retrieval quality issues before deployment.

Available on PyPI.