Step 1
PDF/OCR pass
We extract the normal text layer from PDFs first, and treat image uploads as page images.
Digitise is ready for small samples: upload a PDF or image and see the extraction flow. The production direction is simple: standard PDF/OCR extraction first, then Groq vision checks the page image and corrects what the first pass missed.
We extract the normal text layer from PDFs first, and treat image uploads as page images.
Groq reads the page image and corrects missing, messy, handwritten, or layout-sensitive text.
You get text back in a copyable format without us storing the uploaded file or result.
Use a small sample. The public trial is capped at 3 pages total and remembers trial usage so it cannot be abused.
Digitise returns the visible text as closely as possible, preserving headings, tables, lines, and unclear marks.