AN INDIAN DEV JUST KILLED THE ENTIRE "DOCUMENT PARSING API" INDUSTRY.

PDFs. PowerPoints. Word docs. Images. Videos. Audio. Web pages. One repo eats all of it and spits out clean Markdown ready for your RAG pipeline.

It's called OmniParse. Built by Adithya S Kolavi. 6.8K stars. 537 forks. Sitting on GitHub. Free.

Unstructured.io raised $65M to do this. LlamaParse charges per page. AWS Textract bills you by the request. Azure Document Intelligence wants $1.50 per 1,000 pages forever.

https://x.com/sukh_saroy/status/2059523822441939308?s=46&t=hr2fbvcHJGpvp_SbDstdzg