Edit the real text
inside any PDF.
Not a comment layer on top.
Change the words and keep the layout. Compress to a target size, OCR scans and convert to Word or Excel, right in your browser. No installs, no signup to start.
Open your file
Drop in a PDF or a photo of a document. Scans work too: built-in OCR reads the text off the page.
Say what to change
Type an instruction, or draw a box round the words and edit them in place. Everything else on the page stays exactly where it was.
Download
Your finished PDF, with no installs and no watermark. Save it to your library to come back to it.
Free to use, with 2 free downloads. Then a Week Pass ($6.99) or Pro ($14.99/month). See pricing
Edit PDF
Draw a box to change text, or add text, signatures and stamps.
Sign PDF
Add a signature, date or stamp, by voice.
Protect PDF
Add a password (AES-256 encryption).
Unlock PDF
Remove a password you know from a PDF.
Compare PDF
Diff two PDFs, get a change report.
Compress PDF
Shrink the file, or fit it under a size limit.
OCR PDF
Make a scanned PDF searchable & selectable.
Enhance scan
Sharpen a blurry or faded scan so the text is readable again.
Merge PDF
Combine multiple PDFs into one document.
Split PDF
Separate pages into multiple files (ZIP).
Rotate PDF
Turn all pages 90°, 180° or 270°.
Organize PDF
Reorder or remove pages, across one PDF or several.
Page numbers
Stamp a page number on every page.
Watermark
Stamp text diagonally across every page.
PDF → Word
Extract the text into an editable .docx (OCR for scans).
Word → PDF
Convert a Word document (.docx) to a clean PDF.
PDF → Excel
Pull tables out of a PDF into an editable .xlsx spreadsheet.
Excel → PDF
Convert an Excel spreadsheet (.xlsx) to a clean PDF.
PDF → JPG
Convert each PDF page into a JPG image (ZIP).
JPG → PDF
Combine images into one PDF, a page each.
PowerPoint → PDF
Convert a .pptx slide deck to a clean PDF.
PDF → PowerPoint
Turn a PDF into a .pptx deck, pages become slides.
npm install @nucleushq/marqpdf, or the Python package; webhooks and workflows included.
MCP server for AI agentsGive Claude, Cursor or your own agent PDF tools: remote at /v1/mcp, or local with npx.
Bring your own modelRun extraction, Q&A and the document agent on your Anthropic, OpenAI, Gemini or OpenRouter key.