About
We build tools for documents that are hard to read by machine.
Papers with equations, scanned textbooks, reports that are mostly tables. Doc2X reads the structure of the page first, and everything else it does is built on that.
- What Doc2X does
It translates PDFs into 11 languages and gives back a document that still looks like the original: the same pages, the same tables, the same equations with the same numbers.
It works on scanned PDFs as well as digital ones, and paid plans can export the result as Word, Markdown, LaTeX or Typst.
- How it works
Most tools start from the text they can pull out of a file. Doc2X starts from the layout. It works out the reading order, recognises every formula as LaTeX and every table as rows and cells, and only then translates the prose.
That is why a formula comes back as a formula and a table keeps its columns.
- Who makes it
Doc2X is made by NoEdgeAI LIMITED, a company registered in Hong Kong.
We also run a separate service for mainland China. Accounts, subscriptions and files are not shared between the two.
- Get in touch
Questions, problems and ideas go to [email protected].