01 · Start here
Create an Audio Overview
Follow the exact sequence: process sources, open Studio, choose a format, set language and length, add instructions, then generate.
See the steps →★ Complete Audio Overview HubUpdated July 31, 2026
NotebookLM Audio Overview: Prompts, Formats, Limits & Fixes
Create better Audio Overviews with a step-by-step setup, a compact prompt framework, four-format guidance, current limits and clear troubleshooting paths.
TL;DR — To create a useful NotebookLM Audio Overview, define the listener and outcome, select a focused source set, choose Deep Dive, Brief, Critique or Debate, add custom instructions, then verify important claims while listening. Use the six intent paths below for prompts, formats, limits, failures and the podcast question.
Updated 2026-07-31 · Query-aligned edition. Official Google documentation is separated from editorial recommendations and undocumented behavior. See sources →
What is NotebookLM Audio Overview? It is Google’s AI-generated spoken discussion based on the sources selected in a notebook. “Podcast” is the common nickname; Audio Overview is the official feature name. The best results come from a focused source set, the correct format, clear custom instructions and a verification pass.
Audio is a comprehension layer—not a replacement for the original sources · Important claims still require verification
What do you want to do with Audio Overview?
Six distinct search intents
Start with the task. Each path answers one primary question so the guides reinforce rather than duplicate one another.
02 · Improve the output
Customize the Audio Prompt
Control audience, focus, depth, tone, exclusions and the final takeaway with a compact instruction structure.
Copy prompt examples →03 · Choose the shape
Create Different Audio Formats
Compare the four official formats, then use instructions to create study reviews, executive briefings, interviews, FAQs and storytelling variants.
Compare formats →04 · Check capacity
Audio Overview Limits
Check current daily caps, plan access, language restrictions, length controls, sharing rules and undocumented variables.
See current limits →05 · Diagnose a failure
Audio Overview Not Working
Troubleshoot a missing Generate button, processing failure, stuck generation, missing options, playback trouble or quota message.
Fix by symptom →06 · Understand the name
Podcast vs Audio Overview
Learn why people call it a podcast, what the feature can export, what it cannot publish, and where copyright and fact-checking matter.
See the difference →NotebookLM Audio Overview: quick answers
Most-searched questions
These are the answers people most often need before they generate audio. Use the linked specialist guide when a limit, failure or prompt needs a fuller explanation.
Daily limit
How many Audio Overviews can you make per day?
Current consumer-plan limits are Standard 3/day, Plus 6/day, Pro 20/day, Ultra 20 TB 100/day, and Ultra 30 TB 200/day. Google says daily quotas reset after 24 hours and may change.
Check current limits by plan →Custom instructions
How do you customize an Audio Overview?
Open Audio Overview customization before generation and specify the listener, central question, evidence to emphasize, tone, exclusions and desired takeaway. Instructions steer the discussion; they do not give line-by-line script control.
Use tested instruction templates →Prompt length
Is there a custom-instructions character limit?
Google does not consistently publish a stable public character limit for every account and interface. Keep instructions compact and prioritize audience, focus, evidence and outcome instead of filling the box.
See confirmed vs undocumented limits →Length control
Can you make Audio Overview shorter or longer?
Where the length control is available, you can choose Shorter, Default or Longer. It is a preference rather than an exact duration guarantee, and availability can vary by language, format, account and rollout status.
Understand audio length controls →Hosts
Does every Audio Overview use two hosts?
No. Deep Dive, Critique and Debate use a conversational multi-host structure, while Brief is designed as a compact single-speaker summary. Choose the official format before trying to reshape it with a prompt.
Compare the four official formats →Terminology
Is NotebookLM Audio Overview the Google AI podcast?
“Google AI podcast” and “NotebookLM podcast” are common search terms for Audio Overview. Audio Overview is the official feature name, and the output is not a full podcast publishing system.
Podcast vs Audio Overview →How to create a NotebookLM Audio Overview step by step
Fast path
Open a notebook and confirm the sources finished processing
Audio quality depends on usable source text. Remove failed, duplicate or irrelevant sources before generating.
Open Studio and choose Audio Overview
Select the source subset for this audio instead of automatically using everything in a large notebook.
Choose Deep Dive, Brief, Critique or Debate
The official format sets the basic structure. Do not try to force a Debate prompt into Brief or a one-minute summary into Deep Dive.
Set output language and available length
Shorter, Default and Longer are currently documented for English Audio Overviews. Treat the control as a preference, not a guaranteed runtime.
Add custom instructions
Name the listener, the central question, the evidence to emphasize, what to exclude and what the listener should understand or do at the end.
Generate, listen and verify
Check whether the hosts accurately represent the sources. Correct weak evidence or conflicting documents before simply making the prompt longer.
What NotebookLM Audio Overview formats are available?
The first decision
NotebookLM currently provides four official Audio Overview formats. Custom instructions can shape the audience and presentation, but they do not create a new official mode or guarantee a precise script.
Two hostsDeep Dive
Connects ideas across sources in a fuller conversation. Best for learning, synthesis and explanation.
One speakerThe Brief
Delivers key takeaways from a document in under two minutes. Best for fast orientation.
Two hostsThe Critique
Constructively evaluates an essay, design document, proposal or other draft.
Two hostsThe Debate
Explores competing perspectives in a formal back-and-forth instead of forcing premature agreement.
The five-step Audio Overview workflow
Sources before prompt length
Define the listening outcome
Decide whether the listener needs understanding, a rapid decision brief, critical feedback or competing viewpoints.
Select only relevant sources
A smaller coherent source set usually produces a more useful discussion than a notebook containing every related file.
Choose the format
Use Deep Dive, Brief, Critique or Debate according to the job—not according to which format sounds most entertaining.
Specify audience, focus and ending
Tell the hosts who is listening, what evidence matters, what to exclude and what the final takeaway should accomplish.
Listen while verifying
Open relevant quotes, citations and sources during playback. Regenerate when the framing is weak; correct the source set when the evidence is weak.
Emerging Audio Overview strategies
Beyond one-click summaries
Source-gap audit
Generate audio before the notebook is “finished”
Listen for vague transitions, unsupported confidence or missing counterarguments. These moments often reveal thin, conflicting or poorly organized sources.
Active learning
Pause, predict and question
Use the audio as a first pass, then ask source-grounded questions, create a quiz or use Interactive Mode where available. Passive listening alone is easy to overestimate.
Two-pass production
Orient first, specialize second
Create a broad Brief or Deep Dive to find the important themes, then generate a second audio from a narrower source selection and a more specific audience.
Publishing bridge
Separate synthesis from distribution
NotebookLM creates an audio file, not a podcast series. Fact-check, label AI generation, confirm source rights and use separate editing and hosting tools before publishing.
Read the publishing boundaries →Be careful with “make it very long” prompts. Community workarounds may increase runtime, but longer audio can add repetition and false confidence rather than depth. Improve source selection and segment the topic before forcing length.
When Audio Overview is the wrong output
Use editorial judgment
Exact figures matter. Use a table or written brief when one misstated number could change a decision.
Word-level accuracy matters. Legal, compliance, tax and policy language should be reviewed in the original text.
The evidence is visual. Charts, diagrams, maps and page layouts often lose meaning when converted to audio.
You need precise citations. Audio is useful for orientation, but not a substitute for locating and quoting the source passage.
Three common problems, three better next steps
Do not guess
Too generic
Narrow the source set
Select the few sources that answer one question, then name the audience, evidence focus and desired conclusion.
Improve the prompt →Will not generate
Identify the failing layer
Check edit access, quota, source processing, format availability, browser state and device limitations in that order.
Run the diagnosis →Limit reached
Confirm your actual plan
Google lists separate caps for Standard, Plus, Pro and two Ultra tiers. Daily quotas reset after 24 hours.
Compare current Audio Overview quotas by plan →Important material
Verify while listening
Use background playback while opening relevant quotes, getting citations and querying the source material.
Use a grounded workflow →Audio Overview FAQ
Direct answers
What is NotebookLM Audio Overview?
It is an AI-generated spoken summary grounded in the sources selected in a NotebookLM. It can explain, brief, critique or debate the material depending on the chosen format.
Which format should I use?
Use Deep Dive for connected explanation, Brief for a sub-two-minute summary, Critique for constructive evaluation, and Debate for multiple perspectives.
Can I control what the hosts discuss?
Yes. Select the sources, choose a format and enter a custom prompt that defines the focus, audience and expertise level. You can later use View custom prompt to inspect the instruction associated with the generated artifact.
Can I ask the hosts questions?
Interactive Mode lets you join newly generated Audio Overviews and ask questions with your voice. Google currently documents it as English-only and warns of delays and occasional voice glitches.
How many Audio Overviews can I create?
Google currently lists daily caps from 3 on Standard to 200 on the highest Ultra tier. See the dedicated limits page for the full table and verification date.
Is NotebookLM Audio Overview the same as a podcast?
“Podcast” is the common nickname, but Audio Overview is Google’s official feature name. It creates a downloadable AI-generated audio file; it does not provide an RSS feed, episode management or a publishing workflow.
How do I make an Audio Overview less generic?
Use fewer, more coherent sources and specify the listener, central question, evidence priorities, exclusions and desired takeaway. If the audio stays vague, inspect the source set for missing or contradictory evidence.
Can I make interview, FAQ or storytelling audio?
You can request those presentation styles through custom instructions, usually starting with Deep Dive or Brief. They are prompt-shaped variants, not separate official Audio Overview formats, and the model may not follow an exact script.
Official sources and verification notes
Primary documentation
Google Help
Generate Audio Overview
Formats, prompts, language, playback, sharing, Interactive Mode and known limitations.
Open official help →Google Help
Plans and usage limits
Current daily Audio Overview caps by Standard, Plus, Pro and Ultra plan.
Open official limits →Google Help
Output language
How NotebookLM selects and changes the language used by generated artifacts.
Open language help →Google Blog
Multilingual Audio updates
Background on the rollout of fuller Audio Overviews across more than 80 languages.
Read the announcement →Verified 2026-07-31. Product behavior and quotas can change without a page redesign. Where Google does not publish a fixed value, this guide labels the item as undocumented rather than guessing.
Studio OS · Audio Overview prompt library
Turn two free prompts into a reusable production system
The Studio Collection adds 28 Audio Overview prompts for research, executive briefings, critiques, debates, teaching and Interactive Mode—30 total with the two free templates on this page.
Research Deep Dives · 6
Executive Briefs · 5
Critique Workflows · 5
Evidence Debates · 4
Teaching Audio · 4
Interactive Steering · 4
Next step · Studio OS
Studio OS
Turn a trusted source library into professional briefings, slides, visuals, and presentation-ready assets.
- Reusable, source-grounded workflow architecture
- Complete prompt library with implementation steps
- One-time purchase · permanent access
Privacy and responsible AI use
Before uploading client, student, health, legal or employee information, remove unnecessary identifiers and use an approved organizational environment. Audio makes source content easier to consume; it does not reduce the sensitivity of the underlying material.