mithun50/TreeDex
Tree-based, vectorless document RAG framework. Connect any LLM via URL/API key.
Organizes documents into hierarchical tree structures using font-size detection and PDF outline extraction—eliminating vector embeddings entirely. Supports 15+ LLM providers (OpenAI, Claude, Gemini, local Ollama) plus any OpenAI-compatible endpoint, with identical APIs across Python and Node.js for cross-language index portability. Includes agentic retrieval mode for direct answer generation and automatic orphan-repair for malformed hierarchies.
Available on npm.
Stars
29
Forks
3
Language
TypeScript
License
MIT
Category
Last pushed
Mar 22, 2026
Commits (30d)
0
Dependencies
2
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/mithun50/TreeDex"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
ConardLi/easy-dataset
A powerful tool for creating datasets for LLM fine-tuning 、RAG and Eval
ItzCrazyKns/Vane
Vane is an AI-powered answering engine.
DS4SD/deepsearch-toolkit
Interact with the Deep Search platform for new knowledge explorations and discoveries
xuwei95/ezdata
基于python和llm大模型开发的数据处理和任务调度系统。...
ModelEngine-Group/DataMate
DataMate is an enterprise-level data processing platform designed for model fine-tuning and RAG...