Paperless-ngx vs Hazel

People researching "paperless-ngx vs hazel" are usually trying to answer a single practical question: should documents live inside a dedicated archive, or should they stay as ordinary files in ordinary folders that something tidies automatically? The two tools answer that question differently, which is why a head-to-head comparison is less about features and more about which model fits how you actually work.

Paperless-ngx is a self-hosted document management system, typically run through Docker on a home server, NAS, or spare machine. You feed it scans and PDFs through a consume folder, an email import, or a web upload. It performs OCR on images and non-searchable PDFs, extracts dates, and files each document against metadata such as correspondent, document type, tags, and storage path. Once a document is in the archive, you find it through full-text search in a browser rather than by remembering where you filed it. The originals are preserved alongside the processed versions.

Hazel takes the opposite stance: your folders stay your folders. It runs as a macOS System Settings pane, watches locations you choose, and applies ordered rules when files appear or change. A rule might match a PDF whose text contains a vendor name, rename it with a date pattern, and move it into a client folder. Nothing is imported into a separate database, so every other app on your Mac sees the results immediately. Understanding this split — archive versus in-place automation — is the core of the comparison, and it matters because the wrong model creates friction on every document you handle for years afterward.

Paperless-ngx works as a pipeline. A document lands in the consume directory, and the ingestion worker picks it up, runs OCR where the text layer is missing, generates thumbnails and an archived PDF, then applies matching rules to assign tags, correspondents, and document types. Matching can be literal, regular-expression based, or trained on your prior filing decisions. Everything is indexed for full-text search and exposed through a web interface and REST API. Because it is self-hosted, you own the stack: the database, the storage, the backups, and the updates. That control is the appeal, and the maintenance is the cost.

Hazel works as an event-driven rule engine. You point it at folders, define conditions on attributes such as name, kind, date, source URL, Finder tags, or full PDF text, and attach actions like rename, move, sort into subfolders, run a shell script, or import into another app. Rules are evaluated in order per folder, so ordering and specificity determine the outcome. The logic is explicit and predictable, which is exactly what some people want — and it also means you write and maintain every rule yourself, and each new document category tends to require a new rule.

Sortio sits between these approaches. You describe what you want in plain language — for example, "group these scans by client and year" — and Sortio proposes an organization plan for files that remain in your normal folder structure, with optional renaming and Smart Folders for ongoing automation. It can sort by filename and metadata alone, or read file contents when you turn that on. Content analysis only occurs when you explicitly enable the content sorting toggle. Sortio runs on macOS 10.14+, Windows 10/11, and Linux, backs up files before changes so a sort can be reverted, and offers an offline mode; offline mode processes files locally on your device without cloud connectivity.