The short answer
Add readable sources to a notebook, select Audio Overview in Studio, choose an available format and output language, describe the focus, then generate. Listen with the original sources open and correct your own understanding before sharing.
- Keep sources readable and relevant
- Specify audience and focus in the prompt
- Verify facts against source passages
A source-to-audio walkthrough
0 / 4 steps · Progress is saved in this browser
Add sources
Create a notebook and add documents, text or supported links. Check that uploads are complete; scanned PDFs may need OCR.
Choose the format
Google documents Deep Dive, Brief, Critique and Debate. Pick the one that matches the task rather than using two hosts by default.
Set language and focus
Choose an output language and prompt the topics, audience and level of detail. Available duration controls may depend on language and account.
Generate and verify
Open the output in Studio. Listen for unsupported claims, names and audio glitches; return to the source when something sounds certain but unproven.
Example
A better Audio Overview prompt
“Explain the three findings in these sources to first-year students. Spend most time on the second chart, define technical terms once, and do not add figures not present in the documents.”
When to use another workflow
NotebookLM
Ideal for understanding a collection of sources and hearing a first-pass synthesis.
Editable podcast creator
Prefer this when you need an original script, controlled voices, credits and a final episode for an audience.
Simple text narration
Sufficient when you already have a complete, verified script and only need it read aloud.
Frequently asked questions
Can I choose the language?
Google documents an output-language control for Audio Overviews, with many languages supported.
Can I download the audio?
The current Studio interface provides download options, subject to notebook and source restrictions.
Is it an editable multitrack podcast?
No. Use an audio editor or podcast creator if you need to change individual voice lines and mixing.