OCR.chat

OCR.chat is a web tool for extracting text from documents and then chatting with the result. It accepts images, PDFs, Word documents, and pasted screenshots or text, and it handles OCR in more than 100 languages. The site also presents it as a way to work with tables and math in those documents.

Two reading modes are listed. A fast engine handles printed text, while a premium AI engine reads handwriting, complex layouts, tables, and math. The service can describe what an image shows in plain language, detect language, and turn equations and formulas into LaTeX. It also supports side-by-side review of the extracted text with the original, with low-confidence spans flagged before use. Tables are output as real Markdown or CSV, and extracted content can be copied or downloaded as TXT, Markdown, DOCX, searchable PDF, JSON, or LaTeX. After extraction, users can chat with a document and get answers cited to the page.

Delivery is through the website, with drag-and-drop upload, file browsing, paste input, and a URL option. Multi-page PDFs and batches are welcome, and the page says files are processed privately and deleted automatically. A simple API is also available; one POST request returns clean Markdown or JSON, and the API is metered per page. The site lists PNG, JPG, WEBP, GIF, BMP, TIFF, and multi-page PDF among supported file types.

OCR.chat offers use with no signup for basic text extraction. A free account unlocks the premium AI engine, more pages each month, and OCR history, while paid plans unlock unlimited pages, batch processing, and API use. The page also refers to the service as OCR.chat Tools and includes links for pricing, API, and language selection.