BREAKING
Self-Hosted API Turns Docs to JSON
OCR Plus LLM Pipeline
1
Input PDF/image
↓
2
OCR + vision model
↓
3
LLM structuring
↓
4
JSON / Markdown
Fully Self-Hosted, No Cloud
Switchable OCR Strategies
easyocr
Apache
●
30+ languages
●
Strong in English
llama_vision
Meta
●
Llama 3.2 Vision 90B
●
High accuracy, slow
PII Removal and RAG Prep
MIT-Licensed, Feedback Wanted
AI NEWS BLITZ
An open-source API converts documents into structured JSON or Markdown using OCR and local LLMs.