NotebookLM popularized the source notebook: upload trusted material, then ask questions grounded in those sources. Open Research Notebook follows that workflow, but is open-source, self-hostable, and designed for extensible research tooling such as paper search, custom parsers, tables, and project artifacts.

Those tools are strong at literature discovery and paper Q&A. Open Research Notebook is a project workspace first: sources, chats, notes, generated artifacts, and citation inspection stay attached to the notebook.

It should not. The assistant helps locate passages, compare sources, and draft outputs, but citations stay visible so you can inspect the original material.

No. The notebook can work with PDFs, Markdown, text, DOCX, and web sources. It is optimized for research workflows, but the same source-grounded pattern works for technical notes, market research, policy documents, and project knowledge.

Citations point to the source record and passage used by the answer. For project PDFs and Markdown sources, the goal is to open the source view directly so you can inspect the cited context.

No AI system can guarantee perfection. Open Research Notebook reduces risk by making sources and citations part of the product surface instead of hiding them behind a chat response.

In the default open-source runtime, the app is single-user and runs against your configured database and storage. Hosted deployments can add account isolation, auth, and operational controls in the private cloud overlay.

Your data and files are protected. When you delete, we permanently remove them. No trash, no recovery, delete means delete.

The open-source runtime is single-user by default. Hosted team workspaces, role-based access, and SSO belong in the cloud overlay or downstream deployments.

Zotero import is in development. PDF upload works today. Bibliography sync lands during the pilot.