Open Notebook

Open Notebook is a self-hosted alternative to Google's NotebookLM for researchers, students, and professionals who want AI help with their own research materials. It runs locally or on your server through Docker and is open source under the MIT license. You choose which content the AI can access and which provider processes it.

Local model support includes Ollama, LM Studio, and OpenAI-compatible endpoints. Ollama can handle both chat and embeddings locally, while oMLX supports those tasks on Apple Silicon. You can also connect cloud providers such as OpenAI, Anthropic, and Google. Choosing a cloud provider sends the selected context to that service, so self-hosting the notebook doesn't make every AI interaction local.

The notebook brings PDFs, web pages, Office documents, video, and audio into separate research projects. You can write notes yourself or ask AI to summarize material and extract ideas. Chat uses your research as context and includes source references. Full-text search finds matching words, while vector search helps find related content across your collection.

Podcast generation turns research and notes into audio discussions with customizable voices, speaker profiles, and episode profiles. Model support includes reasoning models such as DeepSeek-R1 and Qwen3. For people who want to adapt the notebook to their work, custom content transformations provide reusable ways to process material, and a REST API supports integrations and automation. The interface supports English, Portuguese, Chinese, Japanese, Russian, and Bengali.