# OmniParse

> Self-hosted OmniParse converts documents, multimedia, and web content into structured markdown optimized for LLM apps like RAG. Open-source solution with Docker.

- **Category:** Document Management
- **Licence:** GPL-3.0
- **Open source:** oss
- **Languages:** Python
- **Platforms:** Docker, pip
- **Tags:** Ocr, LLM, Rag, Fastapi, Transcription, Computer vision, Web Crawler, Gradio, Genai, Document Parsing, Parse Server, Whisper Api, Ingestion Api, Omniparser, Parser Library, Vision Transformer

## Description

OmniParse is a platform for ingesting and parsing unstructured data including documents, images, videos, audio files, and web pages into structured markdown that works well with GenAI applications. It operates locally without external APIs, supports around 20 file types, and provides table extraction, image captioning, audio/video transcription, and web crawling capabilities. The software runs as a FastAPI server with an interactive Gradio UI and requires a GPU with 8-10 GB minimum VRAM due to its use of deep learning models.

## Links

- [Homepage](https://github.com/adithya-s-k/omniparse)
- [Website](https://omniparse.cognitivelab.in/)
- [GitHub Repository](https://github.com/adithya-s-k/omniparse)

---
Source: https://selfhost.directory/project/omniparse