Open Notebook
github.com/lfnovo/open-notebook- Category
- AI Tools
- Rank
- No. 304Tools index
- Type
- APP
- Builder
- lfnovo
- GitHub
- 38.9k stars
- Latest release
- v1.14.0
- Date
About
An open-source, more flexible implementation of Google's NotebookLM for research and Q&A over your sources.
What it does
Open Notebook turns documents, links, audio, video, and pasted text into organized research workspaces. Users can search by keywords or meaning, ask synthesized questions, chat with selected source context, save notes, run repeatable transformations, and produce multi-speaker podcasts.
Why it's ranked here
A strong choice for researchers who value deployment and model control. It combines broad source handling, explicit context selection, multiple search modes, podcast generation, and a REST API. Local model support can keep processing private, while cloud providers remain available when quality or speed matters.
What's good
Users can assign different providers to language, embedding, transcription, and speech tasks. Chat accepts full selected sources, while Ask retrieves relevant chunks for focused synthesis. Keyword and vector search cover precise lookup and conceptual discovery. Custom transformations also support consistent extraction across many documents.
Tradeoffs
Self-hosting brings operational work: Docker Compose runs the application alongside SurrealDB, and source development also needs Python and Node.js. Citations are described as basic. Rate limiting is not implemented, authentication is disabled unless a password is configured, and device synchronization has no built-in solution.
How to use it well
Use it for sustained research projects where sources, conversations, extracted notes, and reusable analyses should remain together. Separate notebooks by topic, choose context deliberately, verify responses through citations, and use transformations for repeated extraction. It does not replace device synchronization, hardened authentication defaults, or comprehensive citation tooling.
Technical notes+
The Python package and dependency surface are declared in pyproject.toml, with FastAPI, LangChain, SurrealDB, Esperanto, Content Core, and Podcast Creator among the dependencies. README.md documents a Next.js and React frontend, Docker Compose deployment, ports 8502 and 5055, and provider configuration through the UI. Makefile exposes development targets for the API, worker, frontend, database, multi-platform Docker images, linting, and release-image testing. docs/SECURITY_REVIEW.md records Fernet encryption for stored credentials, SSRF URL validation, bearer-token authentication, and missing rate limiting. docs/6-TROUBLESHOOTING/faq.md states that the REST API covers UI functionality and that built-in cross-device synchronization is absent.
Observed
- License
- MIT License
- Primary language
- Python 3, requiring Python 3.11 or newer and below 3.13
- Packaging and installation
- Docker Compose is recommended; source installation is also documented
- Interfaces
- Browser-based web UI and full REST API
- Platform support
- Docker installation supports macOS, Windows, and Linux; native Windows guidance is included
- Architecture
- Python and FastAPI backend, Next.js and React frontend, SurrealDB persistence, and a background worker
- Container targets
- Docker builds target Linux AMD64 and ARM64
Read from README.md, Makefile, pyproject.toml, docs/index.md, docs/SECURITY_REVIEW.md, docs/0-START-HERE/index.md, docs/3-USER-GUIDE/index.md, docs/7-DEVELOPMENT/index.md, docs/3-USER-GUIDE/search.md, docs/1-INSTALLATION/index.md, docs/4-AI-PROVIDERS/index.md, docs/5-CONFIGURATION/omlx.md, docs/2-CORE-CONCEPTS/index.md, docs/5-CONFIGURATION/index.md, docs/6-TROUBLESHOOTING/faq.md.
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.