Hey Product Hunt, I built ContextOCR after running into the same problem across my own products. OCR is everywhere. But for AI apps, plain text extraction usually isn’t enough. You still need clean markdown, email rendering, attachment handling, QR/barcode decoding, and output that is easy to pass into an agent or RAG pipeline. ContextOCR is my attempt to make that boring and simple: Send a file → get back AI-ready markdown. It handles PDFs, images, and `.eml` files. For emails, it renders them in a headless browser, follows attachments, and keeps the useful context. It also decodes QR codes and barcodes automatically. I built it first for Passdrop.app and Formslam.com, where it has been working well in production. Now I’m making it available as a standalone API. Curious to hear what people are building with messy documents, screenshots, labels, emails, and attachments.

ContextOCR.dev

OCR for AI: markdown with full context including QR/Barcodes

OCR for AI: markdown with full context including QR/Barcodes