Quick Answer
Google announced new flexible, compute-specific Gemini Notebook limits on 28 August 2026. The rollout starts on 2 September 2026 for consumer accounts on web and mobile.
The new quota refreshes every five hours only until the weekly limit is reached. Google has not published the size of the standard allowance, a compute unit, a token budget, per-feature costs, or the weekly cap. The in-product usage meter is therefore the operational source after rollout.
This change applies to Gemini Notebook at notebook.google.com. It is separate from Gemini Apps and its existing five-hour/weekly limits. The announcement does not put Workspace, Education, or Gemini Notebook Enterprise on the same rollout timetable.
What changes on September 2
Google’s 28 August announcement says the overall limit will account for prompt complexity, chat length, number of sources, and the features used. It also says the notebook can suggest alternative outputs when the first choice surpasses the limit and can defer outputs such as Video Overviews or Slide Decks.
The usage-limits help page supplies the operating rule: the new limits are compute-based, and quota refreshes every five hours until the weekly limit is reached. Its factor list names prompt complexity, models and features, chat length, and specific features. These two official lists are reported separately because Google did not publish a combined rate card.
| Date | What Google published | Boundary to preserve |
|---|---|---|
| 28 August 2026 | Announcement date for the new flexible limits | Not the date the new meter starts |
| Starting 2 September 2026 | Rollout to consumer accounts on web and mobile | A rollout start, not a promise that every account changes at 00:00 |
The scope phrase matters. The work-or-school help page, checked on 30 August 2026, still publishes edition-based feature counts and does not mention the new compute meter, September 2, or a five-hour refresh. The upgrade page separately describes the Google Cloud enterprise path. Until Google publishes a matching change for those account types, this article does not extend the consumer timetable to them.
The new plan table is relative, not absolute
The usage-limits page publishes only relative allowances:
| Plan | Published limit |
|---|---|
| Without a plan | Standard limits |
| AI Plus | Two times higher than standard limits |
| AI Pro | Four times higher than standard limits |
| AI Ultra | 5x or 20x higher than AI Pro depending on the subscription |
Ultra’s multiplier is relative to AI Pro, not to the free standard row. The page does not map the 5x or 20x option to a named storage plan or price.
The following numbers remain unpublished as of 30 August 2026: the standard allowance, the compute unit, a token or credit figure, the cost of one chat or artifact, and the weekly cap. Google also warns that access and limits may change because of testing, experimentation, or availability.
The current feature counts are not the new compute budget
The Google AI Plans table on the upgrade page, checked on 30 August 2026, still says daily quotas reset after 24 hours and monthly quotas after 30 days. It lists this current feature-count card:
| Feature | Standard | Plus | Pro | Ultra (20 TB) | Ultra (30 TB) |
|---|---|---|---|---|---|
| Notebooks | 100/user | 200/user | 500/user | 500/user | 500/user |
| Sources | 50/notebook | 100/notebook | 300/notebook | 500/notebook | 600/notebook |
| Chats | 50/day | 200/day | 500/day | 2.5K/day | 5K/day |
| Audio Overviews | 3/day | 6/day | 20/day | 100/day | 200/day |
| Video Overviews | 3/day | 6/day | 20/day; Cinematic: 2/day | 100/day; Cinematic: 10/day | 200/day; Cinematic: 20/day |
| Reports, Flashcards, Quizzes, Mind Maps | 10/day | 20/day | 100/day | 500/day | 1K/day |
| Deep Research | 10/month | 3/day | 20/day | 75/day | 200/day |
| Data Tables, Infographics, Slide Decks and Revisions | Limited | More limits | High limits | Higher limits | Highest limits |
Those are the counts visible before the announced rollout. The September 2 help notice does not convert them into the new compute allowance or guarantee that they remain the post-rollout artifact budget. Notebook and source inventory are shown per user and per notebook; Google has not said that those inventory caps become five-hour compute limits.
What the Notebook meter shows
The usage-limits help page identifies three places to check:
- Settings → Usage shows AI usage and the weekly limit.
- The chat footer shows remaining usage and a reset time when the account approaches or reaches a limit.
- Studio generation shows an expected AI-usage cost bar before generation.
Gemini Notebook also gets a product-specific Generate later option. Google says it is currently web-only, may take a couple of hours, and sends a notification when enabled. The announcement names Video Overviews and Slide Decks as examples of outputs that can be deferred.
Generate later is a queue, not extra quota. Likewise, an alternative-output suggestion is a way to fit the remaining budget, not evidence of a larger allowance.
For the shared logic of a short refresh window stopping at a weekly ceiling, use the Gemini Apps two-limit guide. The Notebook-specific response is to check its own meter, use Generate later on web when offered, and avoid assuming that another five-hour interval restores an exhausted weekly allowance.
What to do before the rollout
Finish deadline artifacts. If a Video Overview, Slide Deck, or other deliverable must exist before September 2, generate it under the currently published feature-count system. Do not assume a listed daily count becomes an equivalent compute allowance.
Confirm the account type. Consumer web and mobile accounts are the announced scope. Work, school, and enterprise users should keep using the page for their edition until Google publishes a matching transition.
Identify the plan baseline. The new page names no plan, Plus, Pro, and Ultra. Ultra’s multiplier is relative to Pro.
Enable notifications if Generate later is useful. Deferred generation may take hours, so it is not a safe last-minute path.
After rollout, record the product’s own meter. Check Settings → Usage and the Studio cost bar. Do not buy a plan expecting a published prompt count, because Google has not supplied one.
Gemini Notebook and Gemini Apps remain separate meters
The Gemini Apps limits page announced its compute-limit change from 17 May 2026 and uses the same five-hour-until-weekly structure. Its expanded help section says subscribers who reach a limit can continue conversations with Flash-Lite.
The Notebook help page does not publish that Flash-Lite fallback. It instead documents Generate later on web. A shared Google AI plan and similar relative multipliers do not prove that the two products share one usage bar or one fallback.
| Boundary | Gemini Notebook | Gemini Apps |
|---|---|---|
| Product | notebook.google.com | gemini.google.com and Gemini app |
| Compute change | Rolls out from 2 September 2026 | Published from 17 May 2026 |
| Product-specific response | Generate later on web | Flash-Lite conversation fallback for subscribers |
| Heavy examples on the cited page | Video Overviews and Slide Decks; number of sources is a factor | Media generation, Deep Research, advanced models, and higher thinking levels |
Hitting one product’s limit does not, on these pages, establish the state of the other product’s meter.
FAQ
Does a five-hour refresh make Gemini Notebook unlimited each day?
No. Google’s sentence stops the refresh mechanism when the weekly limit is reached. It does not publish the size of either allowance.
Will the current 50-chat standard limit become the new compute limit?
Google has not said that. The current upgrade page still lists 50 chats per day for Standard, while the September 2 page describes a compute-based meter without converting that count.
Does the September 2 rollout include work or school accounts?
The announcement names consumer accounts on web and mobile. The work-or-school page still uses edition-based counts and does not publish the September 2 transition, so this article does not extend the date to those accounts.
Sources
Checked 30 August 2026:
- More compute flexibility in Gemini Notebook — announcement date, consumer rollout, usage factors, alternative outputs, and deferred-generation examples
- Manage your Gemini Notebook usage limits — compute rule, five-hour and weekly limits, relative plan table, meter locations, and Generate later conditions
- Upgrade Gemini Notebook — current Google AI plan feature counts, reset cadence, and the separate Google Cloud enterprise path
- Use Gemini Notebook with a work or school Google account — current Workspace and Education edition table
- Gemini Apps limits and upgrades — separate product, five-hour and weekly mechanism, and expanded Flash-Lite fallback section