Make
HTML to PDF in a real browser, with {{variables}}, repeating invoice rows, page numbers in the footer and bookmarks from headings. Save it as a template and print it by id. Or a public URL, or images.
POST /v1/pdfsPOST /v1/templates
PDF tools & API
Merge, split, stamp, fill and read PDFs. Free in your browser, where files never leave the device, or by API for your code and your agents.
The tools are free. The API is free for 150 documents a month, with plans from $9.
Every tool runs in this tab on the same code as the API. Nothing is uploaded, nothing is kept, and there are no limits beyond your device.
Drop PDFs here or
Files stay on this device. Nothing is uploaded.
No file handy?
This result failed a check, so it isn't offered as a download. If you want to look at it anyway, download it here.
One API for documents: send a file by id, URL, base64 or as the body, and every result comes back as a file whose id goes straight into the next call.
HTML to PDF in a real browser, with {{variables}}, repeating invoice rows, page numbers in the footer and bookmarks from headings. Save it as a template and print it by id. Or a public URL, or images.
POST /v1/pdfsPOST /v1/templatesMerge, split, reorder, rotate and delete pages; stamp text, images, QR codes, watermarks and page numbers.
POST /v1/mergePOST /v1/splitPOST /v1/pagesPOST /v1/stampFill a PDF's own form fields by name, or place values on any PDF by position. Save it as a template; fill by id.
POST /v1/fillPOST /v1/templatesText per page, markdown for LLMs, form fields with their options, and the facts: pages, sizes, encryption.
POST /v1/extractJSON in, files out. Files go in as a fil_… id, a public URL, base64, the raw body, or multipart; results are hosted at a URL for 24 hours, or as long as you ask.
password_required, files[1].pagesX-Gw-Remaining)A free key covers 150 documents a month, no card. Plans from $9 a month for 2,000 documents, up to $79 for 40,000. Get a key · Plans · API reference
# merge two contracts, number the pages, protect the result
curl https://api.pdfedge.com/v1/merge -H "X-API-Key: $KEY" \
-F file=@msa.pdf -F file=@terms.pdf
# → { "id": "fil_8kq2…", "pages": 14, "url": "https://files.pdfedge.com/f/fil_8kq2….pdf" }
curl https://api.pdfedge.com/v1/stamp -H "X-API-Key: $KEY" \
-H "Content-Type: application/json" -d '{
"file": "fil_8kq2…",
"stamps": [{ "type": "page_numbers", "text": "{{page}} of {{pages}}" }]
}'
curl https://api.pdfedge.com/v1/protect -H "X-API-Key: $KEY" \
-H "Content-Type: application/json" -d '{ "file": "fil_9m3x…", "user_password": "open-sesame" }'{
"mcpServers": {
"pdfedge": {
"url": "https://mcp.pdfedge.com",
"headers": { "Authorization": "Bearer YOUR_KEY" }
}
}
}The same operations as MCP tools, so an agent can read a contract, fill a form or bundle a report without glue code. Files go to tools as URLs or ids, never as megabytes of base64 in the chat.
extract_pdf returns markdown for the model, and lists scanned pages it could not readfill_pdf reports unknown and rejected fields instead of failingMost PDF libraries hide a removed page and keep its contents in the file. pdfedge removes it from the bytes.
Merge two forms and every field still works. A name used twice becomes name_2, so filling one never fills the other.
Greek, Cyrillic, Polish, Japanese: text Helvetica cannot draw switches to Noto Sans, or to any Google Fonts family you name.
Bottom-right is the bottom-right the reader sees, on rotated and cropped pages too.
API files are deleted after 24 hours unless you ask to keep them. The tools here never upload at all.
A document is one operation on up to 50 pages. Uploads, reads and templates are free.