Skip to main content
NotebookLM turns your documents into a research assistant that answers questions and generates summaries from sources you choose. The NotebookLM MCP server lets you create notebooks, load them with sources, share them with your team, and generate audio overviews using natural language.

​
What Can It Do?

  • Create and manage notebooks including listing recently viewed ones and deleting in bulk
  • Add sources from text, web URLs, YouTube videos, or Google Drive documents
  • Inspect notebook contents including every source attached to a notebook
  • Share notebooks with teammates by email
  • Generate audio overviews that turn a notebook’s sources into a podcast-style discussion
  • Download generated audio straight into Gumloop storage for use in later steps

​
Where to Use It

Add NotebookLM as a tool to any agent. The agent can then build and populate notebooks conversationally, choosing the right actions based on context. To add an MCP tool to your agent:
  1. Open your agent’s configuration
  2. Click Add tools → Connect an app with MCP
  3. Search for the integration and select it
  4. Authenticate with your account
You can control which tools your agent has access to. After adding an integration, click on it to enable or disable specific tools based on what your agent needs.

​
In Workflows (Via Agent Node)

For automated pipelines, use an Agent Node with NotebookLM tools. This gives you the flexibility of an agent within a deterministic workflow.

​
As a Custom MCP Node

You can also create a standalone MCP node for a specific action. This generates a reusable node that performs one task, useful when you need the same operation repeatedly in workflows.
To create a custom MCP node:
  1. Go to your node library and search for the integration
  2. Click Create a node with AI
  3. Describe the specific action you want (e.g., “Add a web page as a source to a notebook”)
  4. Test the node and save it for reuse
Custom MCP nodes are single-purpose by design. For tasks that require multiple steps or dynamic decision-making, use an agent instead.

​
Authentication

This integration works with NotebookLM Enterprise (also called Gemini Notebook Enterprise), the Google Cloud version of NotebookLM. Consumer NotebookLM at notebooklm.google.com has no public API and cannot be connected.
Sign in with the Google account you use for NotebookLM Enterprise. Access is then governed by the Google Cloud project that NotebookLM Enterprise runs in, so that project needs to be set up first.
Access is granted on the Google Cloud project, not in Gumloop. Each person who connects needs the Cloud NotebookLM User IAM role on that project along with a Gemini Notebook Enterprise license. Connecting the credential still succeeds without them, but tool calls return permission errors. Contact your Google Cloud administrator if you need either one.
What your Google Cloud project needs:
  1. NotebookLM Enterprise set up, with the Discovery Engine API enabled and billing active
  2. The Cloud NotebookLM User IAM role granted to each user who will connect
  3. A Gemini Notebook Enterprise license assigned to each of those users

​
Granting the Cloud NotebookLM User role

A project administrator grants the role once per user, from the Google Cloud console:
1

Open the IAM page

In the Google Cloud console, go to IAM and select the project that NotebookLM Enterprise runs in.
2

Click Grant access

This opens the panel where you add a new principal and assign roles to them.
3

Enter the user

In the New principals field, enter the user identifier. This is typically the email address for a Google Account, a user group, or the identifier for a user in a workforce identity pool.
4

Select the role

In the Select a role list, choose Cloud NotebookLM User, then click Save.
5

Assign a license

In addition to the Cloud NotebookLM User role, each user needs a license for Gemini Notebook Enterprise assigned to them.
Cloud NotebookLM Admin is a separate role for the people who administer Gemini Notebook Enterprise in the project. Working with notebooks through Gumloop only needs Cloud NotebookLM User.
For the full setup flow, including licensing, see Google’s Set up NotebookLM Enterprise guide.

​
Finding your project number and location

Every tool takes a project number and a location. Both appear in the URL you use to open NotebookLM Enterprise:
The project number is the 12-digit number after project= — not the project ID. The location is one of global (default), us, or eu, and must match the region your notebooks live in.
Tell your agent the project number once at the start of a conversation and it will reuse it for every follow-up tool call.

​
FAQ

No. Notebooks created at notebooklm.google.com with a personal Google account aren’t reachable through any API. Only notebooks inside a NotebookLM Enterprise project are available.
Only if your identities are synced into Cloud Identity, so users sign in with a Google identity at notebook.cloud.google.com. Organizations using Workforce Identity Federation sign in at notebook.cloud.google (no .com) as workforce-pool users, and those identities can’t complete a Google sign-in with Gumloop.
Check whether your organization uses VPC Service Controls around Google Cloud APIs. Requests from Gumloop originate outside your network, so a service perimeter can block them even when your roles are correct. Your Cloud administrator can allow them.

​
Available Tools

​
Read Tools

​
Write Tools

With Share Notebook, the people you share with also need the Cloud NotebookLM User role in the same project.
Generate Podcast, Get Podcast Status, and Download Podcast use Google’s standalone Podcast API, which is deprecated and requires separate allowlisting from Google. For audio generation, prefer Create Audio Overview, which works on a notebook’s existing sources.

​
Example Prompts

Use these with your agent or in the Agent Node: List your notebooks:
Create a notebook:
Add a web page as a source:
Add research notes as text:
Inspect a notebook:
Share with a teammate:
Generate an audio overview:
Clean up old sources:

​
Troubleshooting

Agents are smart enough to chain multiple API calls together. For example, asking “Build me a notebook from these three articles” will create the notebook first, then add each source. If results seem off, check the agent’s step-by-step reasoning.

​
Need Help?

Use this integration directly in Claude or Cursor. Connect remotely via the Google NotebookLM MCP server using credentials from your Connectors page.