Upload research papers and PDFs, then ask an AI questions about their content without sending data to the cloud.
Generate multi-speaker audio podcasts from your notes and research materials for learning on the go.
Search across all your documents using both keyword matching and semantic understanding to find relevant information.
Run the entire system locally using Ollama for completely offline AI analysis without any internet dependency.
Requires Docker, SurrealDB, and Ollama (local LLM) to be running; multiple services need coordination.
Open Notebook is a self-hosted, privacy-focused alternative to Google's NotebookLM, an AI-powered research assistant that lets you gather documents and then have conversations with an AI about their contents. The key difference is that you run Open Notebook on your own computer or server, so your research data stays private and you control which AI models power it. The problem it solves is that tools like Google NotebookLM are useful but lock you into one company's AI models and store your data in their cloud. Open Notebook gives you the same core experience, upload PDFs, web pages, audio, and video, then chat with an AI that has read all of them, but with your choice of over 18 AI providers (including local models via Ollama that run entirely offline). It also lets you generate multi-speaker audio podcasts from your research content, search across all your notes using both keyword and semantic search, and access everything via a REST API. You would use this if you do research or study and want an AI assistant grounded in your own documents, but care about privacy, want to avoid cloud subscription costs, or need flexibility in choosing which AI model to use. It runs via Docker and is built with Python on the backend and Next.js with React on the frontend, using SurrealDB as its database.
Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.