NotebookLM MCP

Create notebook

notebook_create

Create an empty NotebookLM notebook and return its ID so you can add sources, start research, or generate content later.

Instructions

Create an empty NotebookLM notebook. Returns the new notebook; keep its id for every later call. Each call creates another notebook, so first look for one with this title in notebook_list. Then fill it: source_add_url, source_add_text or source_add_drive for sources you have, research_start to find some.

Input Schema

Name	Required	Description	Default
`title`	Yes	Title for the new notebook

Output Schema

Name	Required	Description
`status`	Yes
`message`	No	Human-readable outcome or next step
`notebook`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedv0.5.0Oct 20, 2026

changedOutput schema / (root)

Previous value: -nullNew value: +{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "additionalProperties": false,
+  "properties": {
+    "message": {
+      "description": "Human-readable outcome or next step",
+      "type": "string"
+    },
+    "notebook": {
+      "additionalProperties": false,
+      "properties": {
+        "created_at": {
+          "description": "ISO 8601 timestamp",
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "emoji": {
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "id": {
+          "description": "Notebook ID",
+          "type": "string"
+        },
+        "is_shared": {
+          "type": "boolean"
+        },
+        "modified_at": {
+          "description": "ISO 8601 timestamp",
+          "type": [
+            "string",
+            "null"
+          ]
+        },
+        "ownership": {
+          "enum": [
+            "mine",
+            "shared"
+          ],
+          "type": "string"
+        },
+        "sources": {
+          "items": {
+            "additionalProperties": false,
+            "properties": {
+              "id": {
+                "description": "Source ID",
+                "type": "string"
+              },
+              "title": {
+                "type": "string"
+              },
+              "type": {
+                "description": "web_page, youtube, pasted_text, pdf, google_docs, google_slides_sheets, ... or unknown",
+                "type": "string"
+              }
+            },
+            "required": [
+              "id",
+              "title",
+              "type"
+            ],
+            "type": "object"
+          },
+          "type": "array"
+        },
+        "title": {
+          "type": "string"
+        }
+      },
+      "required": [
+        "id",
+        "title",
+        "emoji",
+        "sources",
+        "is_shared",
+        "ownership",
+        "created_at",
+        "modified_at"
+      ],
+      "type": "object"
+    },
+    "status": {
+      "const": "success",
+      "type": "string"
+    }
+  },
+  "required": [
+    "status",
+    "notebook"
+  ],
+  "type": "object"
+}

First observedv0.3.1Aug 6, 2026

TDQS

A4.5/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint=false, idempotentHint=false, and destructiveHint=false; the description reinforces the non-idempotent trait concretely ('Each call creates another notebook') and explains the return value's role ('keep its id for every later call'). It doesn't mention auth requirements, rate limits, or title uniqueness enforcement, so it stops short of fully rich disclosure.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three short sentences: identity, duplicate-avoidance prerequisite, then follow-up actions. The safety-relevant warning is front-loaded ahead of the downstream workflow, and nothing is padding.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With an output schema present, return values need no prose explanation, and annotations cover the safety profile. The description supplies exactly the missing pieces for a non-idempotent creation tool: duplicate-check guidance and the id-retention habit.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Only one parameter and schema coverage is 100%, so the schema already documents 'title' as the notebook title. The description adds no format, length, or uniqueness constraints on the title, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Create an empty NotebookLM notebook') plus the scope qualifier 'empty', which tells the agent this tool allocates a container rather than populating it. This distinguishes it cleanly from source_add_*, research_start, and notebook_rename among the siblings.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly warns that every call creates another notebook and directs the agent to check notebook_list for an existing one with the same title first. It also prescribes the follow-up path by naming source_add_url, source_add_text, source_add_drive, and research_start as the fill options.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.