NotebookLM Enterprise API: Missing Query/Chat Endpoint for RAG Orchestration

Google Cloud Build with AI Agents
,
1

Hello,
We have recently purchased a NotebookLM Enterprise license and are looking to integrate its RAG capabilities programmatically within our organization’s internal tools.
According to the official documentation (projects.locations.notebooks), the available REST API endpoints cover notebook creation, retrieval, listing, sharing, and deletion. However, the get method (Retrieve API) only returns the notebook configuration, system metadata, and a list of attached source links/identifiers—it does not expose the ingested source text content or offer an interface to interact with the underlying vector index.
Our roadblocks:

  1. We cannot find a method to programmatically query the notebook or submit a chat/prompt against the aggregated sources.
  2. The API does not return the extracted or chunked document contents via the management endpoints, meaning we cannot build an external query mechanism over the managed notebook data.

Questions:

  • Is there an unreleased or alpha/beta endpoint (such as a .generate, or .query action) on the projects.locations.notebooks resource that allows querying?
  • If not, what is the recommended architecture for programmatically utilizing NotebookLM’s enterprise RAG pipeline?
  • Is a dedicated query/chat endpoint on the roadmap for the Enterprise API tier?

Any guidance, timeline updates, or workarounds would be highly appreciated.
Thank you!

1 Like