Host Paperless-ngx without installing anything
Drop a scan into a folder or let it collect an inbox; it reads the text, works out the sender and the date, and the document becomes findable.
Starting at €8.99/month
Introduction
Paperless-ngx installed, updated and backed up by us, on a machine in France. Scan and file every document you own. Automatic text recognition, full-text search, tags and correspondents. Your paperwork finally becomes findable. You install nothing and configure no server: the application is online a few minutes after you order, under your own name.
The plan that suits it is App Confort at €8.99/month incl. VAT — 4 GB of memory. Text recognition is the expensive part: OCRmyPDF and Tesseract decompress every page into a full-resolution bitmap before reading it, and several pages can be handled in parallel. Add the full-text search index, the thumbnails produced by Ghostscript, and the periodic retraining of the classifier that learns from your corrections. Between batches the application is very quiet — it is the first five hundred sheets that need the room.
Paperless-ngx does not replace a service so much as a habit: a "to file" folder on the desktop, a cardboard box, or an inbox where invoices go to sleep. There are three ways in — a watched folder your scanner or your phone drops files into, an inbox collected over IMAP whose attachments it extracts according to your rules, and drag-and-drop in the browser. On the way out, the document_exporter tool writes the original files, the OCR-ed PDFs and every piece of metadata as JSON into a folder: your archive stays readable even without Paperless.
What managed hosting changes
Three ways in for documents
A watched folder, where a network scanner or a share from a phone drops its files; an inbox collected over IMAP, with rules that say which attachments to take and which tag to give them; and drag-and-drop in the browser.
OCR, with the original kept
Tesseract reads the text in your scans and Paperless produces a PDF/A with an invisible text layer laid over the image. The original file is not altered: you keep the document exactly as it was scanned, and gain search on top of it.
Filing that learns
Correspondents, document types, tags and storage paths. Rules can be words, regular expressions, or a classifier trained on your own corrections: after a few dozen documents it applies the right tag by itself. The date is read from the content, not from the file name.
Full-text search and batch scanning
The whole pile is searchable by word, including inside the pages. For large batches, separator barcodes split a scanned stack into separate documents, and an archive serial number printed on the sheet ties the physical paper to its record.
Why hand it over rather than install it
Someone emptying boxes of paperwork
Tax notices, bank statements, warranty invoices and payslips go through the scanner once, then come back through a search. No more hunting for the washing machine invoice the night before the repair visit.
A tradesperson whose invoices arrive by email
An IMAP rule collects the supplier inbox, takes the PDF attachments, tags them "purchases" and marks the message as read. By the end of the quarter everything is already filed for the accountant.
An association and its receipts
The treasurer drops in the receipts, the chair reads them, and a grant audit is prepared by filtering on a year and a tag instead of reopening six binders.
An office digitising in bulk
You scan a whole stack in one pass with a barcode sheet between each case: Paperless splits the stack into separate documents and files them. That is what makes digitising an old backlog realistic.
A landlord
Leases, rent receipts, surveys and inventories, one correspondent per tenant and one tag per property. A flat's complete file appears with a single filter.
Hosted in France
Your paperwork is your income, your health and your contracts. It stays in our own datacentre in Nouvelle-Aquitaine, which we operate ourselves, and OCR happens on that same machine, with no outside analysis service.
Frequently asked questions
Three routes, often all three at once. The watched folder: any file dropped there is processed and filed, which lets a network scanner or the mobile app send straight in. Inbox collection: you describe a rule — this sender, this subject, these attachments — and Paperless fetches them by itself. And drag-and-drop in the browser, for one-off documents.
PDFs, JPEG, PNG and TIFF images, and office files when conversion is enabled. On the way out, the document_exporter tool writes the original files, the OCR-ed PDFs and all the metadata as JSON into a folder — tags, correspondents, dates. That export is what guarantees your archive outlives Paperless: nothing is locked in a proprietary format.
OCR runs on the processor: a batch of several hundred pages is handled in the background, and you have to let it work. Handwriting is not recognised — Tesseract reads printed text. Paperless encrypts storage at the server level, not on the client side: it is not a personal-secret vault. And it is not made for signing: a contract that needs a signature goes through a dedicated tool.
App Confort, at €8.99/month incl. VAT. It provides 4 GB of memory where the application asks for 1.5 GB at minimum: the headroom serves the system, not the decor. You can change plan later without reinstalling.
Yes. Paperless-ngx is free software: we host it, we do not own it. And since the export writes your files out in the clear with their metadata beside them, leaving does not mean losing everything.