Docs

Documentation

Deep Notebook turns open-ended questions into structured research runs. Start with a brief, review the plan, monitor the work, and export a cited report with supporting files.

Start here

Quick Start

Start a research session and export a cited report.

Open

Core Concepts

Understand briefs, plans, runs, sources, and reports.

Open

Integrations

Connect private sources and tools for richer research.

Open

Use Cases

See research patterns for markets, companies, and teams.

Open

Reports

Review citations, workspace files, and export options.

Open

Tasks API

Run agents programmatically. Create runs, poll results, and stream events.

Open

Overview

What is Deep Notebook?

Deep Notebook is a deep research workspace for analysts, researchers, and knowledge workers. It helps you turn a question into an editable plan, run an agent across sources and tools, inspect the activity, and produce a report you can reuse.

How it works

A research session starts with your brief, files, connected sources, and model choice. Deep Notebook generates a research plan, lets you edit the steps, then executes the run while showing sources, tool calls, intermediate files, and final report output.

Key concepts

Research sessions

Workspaces for one focused investigation, from brief to final report.

Research plans

Editable steps that define how the agent should gather, compare, synthesize, and write.

Connected sources

Files, apps, repositories, messages, documents, and other context the researcher can use when it is relevant to the session.

Reports and citations

Source-backed deliverables that can be copied, downloaded as Markdown, or exported with generated workspace files.

Getting help

Questions or feedback? Reach us at madhav@autocomputer.ai.