Official
Daily quotas reset after 24 hours
Google does not describe one universal local-midnight reset for every user.
TL;DR — Google currently lists 3 Audio Overviews per day on Standard, 6 on Plus, 20 on Pro, 100 on Ultra 20 TB and 200 on Ultra 30 TB. Daily quotas reset after 24 hours and are subject to change.
Last verified 2026-07-28. Official numbers are separated from observed or undocumented behavior. See primary sources →
Current limits referenceVerified 2026-07-28 · RC14.54
NotebookLM Audio Overview Limits
The current daily caps by plan, the 24-hour reset rule, format and language constraints, Interactive Mode and sharing limits—and the values Google still does not publish.
Direct answer: Audio Overview limits are plan-based. Google currently lists 3/day Standard, 6/day Plus, 20/day Pro, 100/day Ultra 20 TB, and 200/day Ultra 30 TB. Daily quotas reset after 24 hours. These limits are explicitly marked subject to change.
Do not infer a midnight reset · Automatic first artifacts may not count · Sharing a notebook does not change a collaborator’s limit
Audio Overview daily limits by plan
Official Google table
Standard
3
per day
Plus
6
per day
Pro
20
per day
Ultra · 20 TB
100
per day
Ultra · 30 TB
200
per day
| Plan | Audio Overviews | How to confirm | Important note |
|---|---|---|---|
| Gemini Notebook Standard | 3/day | Free consumer account or qualifying access | Do not assume all free-looking accounts have identical organizational policies |
| Gemini Notebook in Plus | 6/day | Check the distinct plan badge near the profile image | Plan activation problems should be diagnosed separately from quota |
| Gemini Notebook in Pro | 20/day | Check plan badge and Google AI plan | Other Studio artifacts have separate quotas |
| Ultra 20 TB | 100/day | Confirm the specific Ultra storage tier | Not the same as the 30 TB tier |
| Ultra 30 TB | 200/day | Confirm the specific Ultra storage tier | Highest published personal-account cap |
Google also says an artifact automatically generated when sources are first added may be generated once without contributing to limits.
How the 24-hour reset works
Not “every midnight”
Practical implication
Record when the cap appeared
If you hit the limit at 4:00 p.m., do not assume access must return at 12:01 a.m.
Collaboration
Limits remain per collaborator
Sharing a notebook does not transfer the owner’s higher limit to another collaborator.
Account check
Confirm the plan badge
Google recommends checking for the distinct badge beside the user profile image.
Format, language and length constraints
Official feature limits
| Item | Current documented status | What it means |
|---|---|---|
| Formats | 4: Deep Dive, Brief, Critique, Debate | Each format performs a different job; it is not merely a style choice |
| The Brief | Single speaker, under two minutes | It is intentionally short, not a failed Deep Dive |
| Output languages | 80+ | Gemini Notebook uses the preferred Google Account language as the default |
| Length choices | Shorter, Default, Longer | Google’s current help page labels the length choices English-only |
| Interactive Mode | English-only | Only newly generated Audio Overviews can use it |
| Generation time | “A couple of minutes” | No guaranteed completion time is published |
| Mobile | May have feature limitations | Use desktop to isolate whether a missing control is mobile-specific |
Sharing and download limits
Access matters
Public links
Public notebook sharing is enabled only for consumer accounts. It is currently disabled for Workspace Enterprise and Education accounts.
Who controls access
Only notebook owners and editors can make generated audio public and manage its access level.
Deleting breaks the link
Once an Audio Overview is deleted, its previously generated share link no longer works.
Shared users cannot join
Recipients can listen to the original shared audio but cannot interact with the hosts through the link.
Downloaded audio can be shared
The web experience includes a Download action; Google does not promise a permanent fixed file specification in the help page.
Voice interaction storage
Google says voice and transcribed interactions with the hosts are not stored or shared.
Official, observed and undocumented are different
Do not turn guesses into limits
Safe to quote with a date
- Daily plan caps
- 24-hour reset statement
- Four formats
- 80+ languages
- Interactive and sharing restrictions
Requires a documented test
- Generation time by format
- Typical duration by length
- Quality at different source counts
- Device-specific behavior
- Rollout differences
Do not invent a number
- Guaranteed maximum duration
- Maximum custom-prompt length
- Exact voice model version
- Internal compute or memory
- Detailed rate-limit mechanics
What Google does not currently publish
A research opportunity
| Question | Status | How to report it responsibly |
|---|---|---|
| Maximum Audio Overview duration | Undocumented | Report measured examples by format, language and date—not a universal cap |
| Maximum custom-prompt length | Undocumented | Test progressively and label the interface behavior observed |
| Generation time by format | No fixed guarantee | Use identical sources and record median plus range |
| Quality threshold by source count | Undocumented | Compare the same task across coherent source sets |
| Cooldown after quota errors | Only 24-hour reset is documented | Record timestamps; do not claim a universal rolling formula without evidence |
| Exact audio model and voices | Not publicly specified | Avoid unsupported model-name claims |
A reproducible Audio Overview benchmark plan
Build evidence over time
Use one fixed source pack
Keep source content constant so the test isolates format, language or length.
Record plan and interface
Capture plan badge, account type, device, browser and generation date.
Change one variable
Test format, language, length or source count separately.
Measure generation and output
Record time to completion, audio duration, failures and source-grounding quality.
Publish the uncertainty
Report ranges and exceptions, then add a Last verified date to every table.
Audio Overview limits FAQ
Direct answers
How many Audio Overviews can free users create?
Google currently lists 3 per day for Gemini Notebook Standard.
Does the daily limit reset at midnight?
Google says daily quotas reset after 24 hours. It does not state one universal midnight reset.
Does sharing a notebook share the owner’s quota?
No. Google says sharing a notebook does not change limits for any collaborator.
How long can an Audio Overview be?
Google does not publish a guaranteed maximum duration. The Brief is documented as under two minutes; other output duration varies.
Are Longer Audio Overviews available in every language?
The current official help page labels Shorter, Default and Longer length choices as English-only, even though Audio Overview supports 80+ output languages.
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 Gemini Notebook 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 2026-07-28. 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.