A clearer view of AI.
Subscribe

WEEKLY NEWSLETTER

The AI Insider

A weekly round-up of the AI tools worth your time, and the hype to ignore.

Unsubscribe at any time using the link in each newsletter.

THE INDEPENDENT AI.

HYPE OR FACT? · CHECKED 4 NOVEMBER 2026

Gemini Notebook API: an official API and MCP server for personal accounts

OUR VERDICT

Fact

Our original verdict: “Fact, but early and narrow. Google really has opened a free Gemini Notebook API and MCP server for personal accounts, but for now it can only create notebooks and add sources: it can't ask questions or make Audio Overviews yet.”

How we judge

THE CLAIM

“Google's developer docs: 'The Gemini Notebook API is a REST interface for programmatically creating and managing notebooks and sources in Gemini Notebook', with an MCP server, in an early access Public Preview for personal Google Accounts.”

Google for Developers · Original post ↗

What’s actually true

  • It's real and official. Google's developer site documents a REST API at notebook.googleapis.com/v1 and a remote MCP server at notebook.googleapis.com/mcp. It's labelled an 'early access Public Preview' (pre-GA), so it can change.
  • It's new, and different from the existing Enterprise API. Until now the only official Gemini Notebook (formerly NotebookLM) API was the Enterprise one on Google Cloud, which needs paid Enterprise licences. This one works with ordinary personal Google Accounts. Google Workspace accounts aren't supported yet.
  • What it does today, by Google's own docs: create, list, rename and read notebooks, and add sources (web pages, YouTube videos, plain text or Markdown), then check when a source has finished processing. The MCP server offers the same seven actions to tools such as Claude Code and Cursor.
  • What it can't do yet: ask a notebook questions, make Audio or Video Overviews, slide decks or quizzes, upload files such as PDFs, add Google Drive files, or delete anything. Google's roadmap lists most of these as coming, with no dates.
  • Price and limits (Google's figures): free during the preview, with rate limits (for example 20 new sources a minute per user) and your usual plan limits: 100 notebooks and 50 sources per notebook on the free plan, more on paid Google AI plans.
  • Setup is technical: you need a Google Cloud project, the gcloud command-line tool and an OAuth sign-in with a special 'notebook' scope, and you must have signed in to Gemini Notebook once. Access tokens last an hour.
  • Google didn't publish a launch post that we could find. The docs pages are dated 29 October to 2 November 2026, and the TLDR AI newsletter picked it up on 4 November.
  • We haven't tested it hands-on.

Sources

Cost

Free during the Public Preview (Google's statement). You need a Google Cloud project; notebook and source limits follow your Gemini Notebook plan (free, or Google AI Plus, Pro or Ultra).

USE IT IF

Developers and tinkerers who want an AI agent to file research into Gemini Notebook automatically, for example adding every competitor page or YouTube video you find to one notebook.

SKIP IT IF

Anyone who wants an agent to query notebooks or make Audio Overviews by API (not possible yet), and Google Workspace business accounts (not supported yet).

How we knowChecked on 4 November 2026. Verdicts: Fact = the claim holds up; Partly fact = real, but with caveats or overstated parts; Hype = the headline claim isn't supported. Products change quickly, so check the maker's site before you buy. Our methods & sources ↗