Self-host Paperless-AI next
v2026.10.02AI document tagging for Paperless-ngx with OCR rescue
- Stars
- 264
- RAM
- ~512 MB
- Port
- 3000 (Web UI)
- Requires
- AI provider (OpenAI, Ollama, Azure, or custom), Paperless-NGX
admonstrator/zettelrobbe
What is Paperless-AI next?
Zettelrobbe is a fork of Paperless-AI that automatically classifies documents in Paperless-ngx using AI. It connects to a Paperless-ngx instance and uses a chosen AI provider — OpenAI, Ollama, Azure OpenAI, or any OpenAI-compatible endpoint — to analyze new documents and assign tags, correspondents, document types, titles, and custom fields. Key capabilities include an OCR rescue queue that re-processes poorly scanned documents using Mistral OCR or local vision models, a duplicates page that finds and merges similar tags and correspondents with AI assistance, a history page with one-click restore of original metadata, manual tagging for on-demand processing, and a simplify-tags feature to consolidate tag vocabularies. The web UI includes MFA login, OIDC single sign-on support, server-side pagination for large archives, and a guided 7-step installer. It runs as a single Docker image (~500-700 MB) with a bundled SQLite database for history and user accounts, and requires a running Paperless-ngx instance.
Should you self-host Paperless-AI next?
Paperless-AI next scores 49/100 on selfhost.directory (how scoring works) and ranks #31 of 46 in Document Management. It has 264 GitHub stars.
It shipped 21 stable releases in the past year. The latest, v2026.10.02, came out on 22 October 2026. New versions land roughly every 8 days.
Setup is moderate: expect some configuration beyond a single command. An official Docker image is available. It is MIT-licensed and written mostly in JavaScript.
Its closest alternative on selfhost.directory is PaperCortex (35/100, 31 stars). Paperless-AI next scores higher overall. See Paperless-AI next vs PaperCortex side by side.
Features
- Automatic AI document tagging
- Custom fields extraction
- Duplicate detection and merging
- External API enrichment
- Guided installer
- History and one-click restore
- Local AI operation
- Manual tagging
- MFA login
- Multi-Provider AI Support
- OCR rescue queue
- OIDC Single Sign-On
- Rate Limiting
- Server-side pagination and tag caching
- Simplify tags
Paperless-AI next README
FAQ
Does this support a specific niche AI provider?
If it's OpenAI-compatible, it probably works. Otherwise, open an issue and the author may implement it.
Why should I use this instead of the original Paperless-AI?
If the original works for you, stay there. Zettelrobbe offers better performance, improved AI and OCR, and a more polished experience.
What if I find a bug?
The author welcomes pull requests from all contributors and is happy to learn from code reviews.
Why was RAG (semantic search/chat) support removed?
The RAG subsystem required a secondary Python container with ChromaDB, adding massive image sizes, high RAM requirements, and architectural overhead. The project now focuses on core document tagging.
Is this project stable if it's vibe-coded?
The author uses it daily for actual documents and has a strong incentive to keep it stable. Code is reviewed and tested before merging.
More in Document Management
Paperless-GPTOpen-source, self-hosted AI companion for paperless-ngx: suggests titles, tags, correspondents and dates, and runs LLM vision OCR for searchable scans.
paperless-iqPaperless IQ is a self-hosted AI add-on for Paperless-NGX adding metadata tagging, RAG search, vision analysis, and approval workflow — supports Ollama, Anthropic, OpenAI
Paperless-ngxSelf-hosted open-source document management system that turns paper documents into a searchable digital archive with OCR, AI features, and multi-user support
papermarkAn open-source, self-hosted DocSend alternative: share documents through custom links, add your own domain and branding, and track views with built-in analytics.
PapermergeOpen-source self-hosted app for organizing scanned files using OCR, search, tags, folders, and a desktop-style web interface.
PapraThe minimalistic document archiving platform.
pdfmanagerSelf-hosted PDF manager offering annotations, nested collections, OIDC authentication, multi-user libraries, sharing, full-text search, and OPDS access for e-readers.
pdfscanexplorerSelf-hosted web app for previewing, renaming, and organizing scanned PDFs using OCR and optional AI, lightweight for NAS or server.
Pen2PDFSelf-hosted productivity suite with AI document extraction, timetable, todos, notes, whiteboard, and multi-model AI assistant. Runs on Node.js with MongoDB.
PetrichorPetrichor is an open-source, self-hosted platform turning Markdown into semantic wikis with agentic RAG for traceable, evidence-backed answers.
PharosSelf-hosted, open-source RAG system with multi-format document ingestion, hybrid retrieval, enterprise ACL, and dual HTTP + MCP exits for team knowledge bases.
PrivateBinA minimalist, open source online pastebin where the server has zero knowledge of pasted data. Data is encrypted/decrypted in the browser using 256 bits AES.
0 comments