NotebookLM API: Access Options and Workflow Limits
Google has not released a public API for Gemini Notebook (formerly NotebookLM). This guide explains what programmatic access currently exists, how to distinguish the product interface from the underlying model APIs, and how fenbs keeps selected research sources and decisions available outside any single notebook.
Updated 4 min read
Gemini Notebook — the research tool Google previously called NotebookLM — does not have a public API as of October 2026. You cannot use an API key to create notebooks, add sources, or query them programmatically. The product is a web interface, and Google has not opened a supported way for code or other tools to call it directly.
If you need to work with selected sources and decisions across multiple AI apps, fenbs gives you a portable record any connected assistant can read. You save the verified sources, key conclusions, and next steps there; the record stays accessible even when a specific product has no API.
What "NotebookLM API" most likely means
When people search for a NotebookLM API, they usually want one of three things:
- A way to create or populate notebooks with code, so sources can be added automatically.
- A way to query a notebook's sources programmatically, getting answers without the web interface.
- A way to read a notebook's audio overview or summary from another app.
None of these are currently available through a documented, supported API. Google's developer documentation does not include a Gemini Notebook API. Before building a workflow that assumes one, check Google AI's developer site for the latest status.
What Google does provide
Google provides two separate things that are sometimes confused with a Gemini Notebook API:
The Gemini API gives you programmatic access to Gemini models. You can send documents and ask questions about them, which overlaps with some Gemini Notebook use cases. However, it is not the same product: you are calling the model directly without the notebook's source management, grounding, or audio overview features.
Google AI Studio is a web interface for the Gemini API. It is not an API itself, but it lets you explore what the API can do before building with it.
If your goal is to ask questions about a set of documents in code, the Gemini API is the closest official option. Read Google AI's developer documentation for current models, context window limits, and pricing before choosing an approach.
Community workarounds and their risks
Community projects have attempted to automate Gemini Notebook by using browser automation or unofficial API calls observed in network traffic. These approaches work until Google changes the interface or restricts the endpoints, which it can do at any time because the interface is not intended for programmatic access.
Before using a community tool that wraps an unofficial interface, consider: if it breaks after an update, does your workflow have a fallback? If the tool is no longer maintained, will you be able to fix it? For workflows you depend on, unofficial wrappers add fragility rather than reducing it.
Distinguishing the product from the model
A common confusion is between Gemini Notebook as a product and Gemini as a model. They are different:
| Gemini Notebook | Gemini API | |
|---|---|---|
| What it is | A web product for research and source management | An API for calling Gemini models in code |
| API available? | No public API | Yes, documented at Google AI |
| Source management | Built-in (add PDFs, web pages, YouTube videos) | You manage context in your own code |
| Audio overviews | Yes | No |
| Grounding | Notebook sources | Google Search, your own data via context |
| Use case | Research with a curated source set, via browser | Programmatic document analysis and generation |
What about the Gemini Notebook export options?
Gemini Notebook does let you copy or share notebook content, but these are manual actions in the web interface, not a programmable API. You can copy a generated overview, share a notebook link with collaborators, or export an audio overview as a file. None of these let an external tool read or write notebooks automatically. They are useful for one-off sharing, not for workflows that need to pull data on a schedule or in response to an event in another system.
Keeping research portable without a Gemini Notebook API
The most practical way to make Gemini Notebook research portable is to maintain a parallel record outside the notebook. Save the sources you have verified, the conclusions they support, and the next steps you identified in a place another AI can read.
This is especially useful when you switch from Gemini Notebook to a different assistant for the next step. The assistant does not need notebook access if the key sources and decisions are already in a shared record it can see.
Keep research decisions available to every connected assistant
Save verified sources in fenbs Links and listings, key conclusions in Decisions and rules, and next steps as tasks on the board. Any AI app you connect over MCP reads the same record. When you switch from Gemini Notebook to Claude, Cursor, or another tool, the context is already there. Free to start.
Create a free accountFree for one person · No credit card · Works with Claude, ChatGPT, Cursor and more
How to do this with fenbs
- Create a free account and add a project for the research area.
- As you work in Gemini Notebook, save the sources you have verified in Links and listings with a note on what each one confirms.
- Write your settled conclusions in Decisions and rules so any connected assistant treats them as agreed, not open questions.
- Add follow-up tasks and open questions to the kanban board.
- Connect your AI apps from Integrations. When you start a new session in any supported tool, ask it to read the project context. It starts from your verified sources and decisions without needing Gemini Notebook access.