Private by design. Clear about its limits.
Your invoices never leave your browser.
We can't see them, structurally: your PDF is read and split locally. Turn a merged invoice PDF into named files, review the page ranges, then download a ZIP.
Local processing · Free up to 10 documents per ZIP
Split your invoice PDF
One PDF · up to 50 MB / 500 pages · no upload · no OCR
What you get
Financial documents stay local
PDF reading, boundary detection, and ZIP creation happen in your browser. No document upload, account, or remote AI service.
Honest, inspectable detection
Choose invoice numbers, fixed page counts, or a literal keyword. Review every range. Pages without text and ambiguous invoice numbers are flagged.
Named files, ready to file
Download individual PDFs together in one ZIP. Invoice numbers become safe filenames; duplicate names receive numbered suffixes.
Pricing
Invoice Splitter
Split a merged invoice PDF into named files — right in your browser.
Planned paid tier — not available yet. Unlimited documents, custom filename patterns, and saved templates are coming later. Leave your email below for launch news.
Hear when the $29 tier is ready
Planned: unlimited documents per ZIP, custom filename patterns, and saved templates. These features are not available yet. The free tool includes the first 10 documents per ZIP.
By submitting you agree we may email you about this product. Privacy policy. We collect only what's listed there, and it stays with Avyth Works LLC.
Questions
Can I split scanned PDFs?
Only PDFs with a usable text layer can pass the text check. Every page without selectable text is flagged. If more than 20% of pages have no text, all presets block export. OCR is not supported in v1: add a text layer elsewhere first. A text layer alone does not guarantee accurate boundaries; always review.
Is my data safe?
PDF bytes, extracted text, invoice numbers, and filenames stay in browser memory. The tool does not send them to a server or save them in browser storage. Close the tab to release the working session. Only the optional email form submits your email address.
How does invoice detection work?
A fixed, case-insensitive pattern looks for Invoice #, Invoice No (with an optional period), or Invoice Number, followed by an optional separator and a non-whitespace value ending at whitespace, a comma, or a semicolon. A different number starts a new document; the same number continues it. Multiple distinct numbers on one page block invoice-mode export. There is no fuzzy matching or user-supplied regex.
What happens to pages with no boundary?
They attach to the previous document. Pages before the first detected document are flagged and block export, so they are not silently lost. Try a suitable keyword or Every N pages. Keyword matching ignores case and starts a new document on every matching page, including repeated headers.
What is free, and what is coming later?
Review all detected documents for free; each ZIP includes only the first 10. If more are detected, a visible notice states exactly how many are omitted and the table marks them. The planned $29 one-time tier would add unlimited documents, custom filename patterns, and saved templates. It is not available to buy or use yet.
What are the file and output limits?
One merged PDF, at most 50 MB and 500 pages. Encrypted or damaged PDFs may not work. Large or complex PDFs can exceed browser memory. Filenames use ASCII letters, digits, hyphens and underscores, capped at 200 characters including .pdf; collisions receive -2, -3, and so on. Splitting copies pages and is not a redaction or PDF sanitization service.