Paperless-ngx Hosting

How Paperless-ngx works

Paperless-ngx receives documents through uploads, a watched consume directory, supported mail ingestion, or its API. It converts supported files into archival records, extracts text through OCR when needed, and keeps the original document together with searchable metadata. The web interface then provides filters, full-text search, thumbnails, and document previews.

Users can assign correspondents, document types, tags, dates, storage paths, and custom fields. Matching rules and workflows can apply metadata or actions during processing, while saved views create repeatable queues for records that need review. The archive remains useful only when the originals, generated files, database, and search index stay consistent.

Key Paperless-ngx features

OCR makes scanned pages searchable, and the application can process common image and office-document inputs through supported parsers. Tags and correspondents organise records from different perspectives, while document types separate invoices, contracts, statements, receipts, and other material.

The interface supports bulk editing, custom fields, notes, permissions, and workflows. An API enables scanners, scripts, and other systems to add or retrieve documents. Search combines text with metadata filters, helping users recover a record without remembering its original file name.

Who uses Paperless-ngx

Households use Paperless-ngx for bills, warranties, tax records, and personal correspondence. Offices can archive supplier documents, contracts, policies, and administrative forms, while small organisations use permissions and workflows to keep shared records organised.

The application is not a substitute for a legal retention policy or an independent backup. OCR results can contain errors, so important names, dates, and amounts should be reviewed when accuracy matters.

Self-hosting Paperless-ngx: requirements and cost

Storage is the primary long-term constraint because the installation keeps original files, generated archival versions, previews, thumbnails, OCR data, and database records. CPU and memory demand rise during imports, OCR, classification, and bulk processing. The catalogue stack includes PostgreSQL and Valkey; PostgreSQL is provisioned with the application but remains unmanaged.

Paperless-ngx has no separate software charge, and the database-adjusted recommendation is Plan 4 at €20 with 100 GB for the complete stack. Storage is the real limit here — 50 GB on Plan 3 is a realistic starting point only before the database step-up, while 100 GB on Plan 4 better suits a growing archive. AvaHost also includes one-click deployment, a custom domain, automated HTTPS, automatic application updates, and scheduled backups.