Paperless-AI

Paperless-AI is an extension for Paperless-ngx users who want to sort incoming documents automatically and ask questions about their archive. The project is no longer maintained.

• Automatic classification: It detects new documents and assigns titles, tags, document types and correspondents based on their contents.
• Archive search and chat: You can search by meaning and ask questions about document contents, such as the date of a rental agreement or the amount on an electricity bill.
• Processing rules: You can limit which documents receive AI analysis and define the tags it applies.
• Manual tagging: A browser interface lets you request AI tagging for selected documents, including files you want to review before processing.

The extension runs in Docker alongside an existing Paperless-ngx instance. It supports Ollama with Mistral, Llama, Phi-3 and Gemma-2, plus OpenAI-compatible backends such as LiteLLM, vLLM and FastChat. A local backend runs model inference on your own hardware. Cloud providers, including OpenAI, DeepSeek, OpenRouter and Gemini, process the document content sent to them remotely.

Paperless-AI is open source under the MIT license.