adithya-s-k/omniparse

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

7815 stars · 667 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

OmniParse is a self-hosted ingestion and parsing platform that converts unstructured data (documents, images, audio, video, web pages) into clean, structured markdown optimized for LLM applications. It runs as a local API server with a Gradio UI, supporting ~20 file types with OCR, table extraction, transcription, and web crawling.

Use cases

• parse resumes from pdfs
• convert documents to markdown for RAG
• transcribe audio and video files locally
• extract tables from images and documents
• crawl web pages into clean markdown for LLM training
• prepare unstructured data for fine-tuning
• self-hosted document ingestion pipeline

When to choose

• you need fully local, private data parsing with no external APIs
• you want one tool handling documents, multimedia, and web content
• you're building RAG or fine-tuning pipelines and need LLM-friendly structured output
• you can run on Linux with optional GPU acceleration

When to avoid

• you need Windows or macOS native support (server is Linux-only)
• you only need simple PDF text extraction without multimedia or web crawling
• you require a managed cloud service rather than self-hosting
• your project needs a permissive license (GPL-3.0)

Facets

service · maturity active

ocr parser web-scraping speech-recognition rag etl api-framework nlp large-language-models developer-tools pdf crawlers self-hosted python data-ingestion document-parsing multimedia-transcription genai-preprocessing markdown-conversion gradio-ui data-engineering natural-language-processing retrieval-augmented-generation linux docker web-server gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="adithya-s-k/omniparse")