Google Switches Gemini Notebook to Compute-Based Usage Limits Starting September 2
Google is switching Gemini Notebook from fixed daily feature limits to compute-based usage limits that factor in prompt complexity, chat length, and sources used. The change, mirroring a May update to the Gemini app, rolls out to consumer accounts on September 2, 2026.
Google is replacing Gemini Notebook's fixed daily generation limits with a compute-based usage system starting September 2, 2026. The change follows a similar shift made to the Gemini app in May, and it alters how the notebook-taking and research tool measures and communicates usage caps to subscribers.
What's changing
Under the current system, Gemini Notebook allocates a set number of feature generations per day based on a user's Google AI subscription tier. The new system removes that fixed daily count and instead calculates usage based on actual compute consumption. According to Google, the overall usage limit will "factor in prompt complexity, chat length, number of sources, and the features you use."
Limits will also refresh every five hours instead of resetting once per day, giving users more frequent access windows rather than a single daily allotment.
New usage indicators
Google is adding visual tools to help users track consumption in real time:
- A status message will appear below the chat window, such as "You're almost at your AI usage limit. Limit resets at 3:00 PM," or "Limit reached. All features are available after 3:00 PM."
- Studio generations will show an expected-cost bar: the more it's filled, according to Google, the higher the expected compute cost of that generation.
- Notebook will suggest alternative outputs if a user's first-choice generation would exceed their remaining limit.
A new "Generate later" option lets users defer compute-intensive outputs — such as Video Overviews or Slide Decks — for automatic generation once their usage limit resets. This deferral feature is launching first on web.
Tier-based multipliers
Rather than eliminating tier differentiation, Google is converting it into a multiplier on top of the compute-based baseline:
- Free/no plan: standard limits
- Google AI Plus: 2x standard limits
- Google AI Pro: 4x standard limits
- Google AI Ultra: 5x or 20x higher than AI Pro, depending on the specific Ultra subscription
The rollout begins September 2, 2026, for consumer accounts on both web and mobile.
What this means
This is a metering change, not a model or product launch — Gemini Notebook itself isn't getting a new underlying model. The shift mirrors a broader industry pattern of moving away from flat per-day quotas toward usage metering tied to actual inference cost, which better reflects the reality that a short summary and a multi-source Video Overview consume vastly different amounts of compute.
For heavy users, the practical effect is uncertain until Google publishes concrete numbers on what "standard limits" actually translate to in generations or compute-units — a detail absent from this announcement. The five-hour refresh cycle is likely a net positive for casual users who previously exhausted a daily cap early and had to wait until the next day, but it could tighten availability for users who cluster heavy usage into short windows, such as running multiple Video Overviews or Slide Deck generations back-to-back. The new "Generate later" queueing feature suggests Google anticipates this friction and is building deferral mechanics to manage user expectations rather than simply throttling access.
Related Articles
Grok Bot agent gets its own @mail.grokbot.com email address, rolling out to users now
Grok Bot, the agent available on iPhone, iPad and Mac, now has its own email address ending in @mail.grokbot.com. According to the announcement on X, the agent can use it to sign up for services, contact businesses and schedule time with people. The rollout began October 9, 2026.
Anthropic launches OSS Scanner: free, model-generated security scans for opt-in open-source projects
Anthropic has launched OSS Scanner, an opt-in service that gives open-source projects periodic vulnerability scans at no cost, run by its strongest models including Claude Mythos. Reports are fully model-generated with no human review or triage, so Anthropic warns some may be incorrect or invalid.
Anthropic adds Dashboards and Motion betas to Claude; Docs, Slides and Design exit beta for all plans
Anthropic launched two beta features for Claude: Dashboards, which builds auto-updating dashboards from connected data sources, and Motion, which generates animated explainer videos exportable as MP4. Docs, Slides and Design leave beta and now work on every plan, including free accounts.
Google releases free AI Edge Foresight, a Mac note-taking app that transcribes meetings fully offline
Google has released AI Edge Foresight, a free experimental macOS app that transcribes meetings and audio files entirely on-device. It is optimized for Apple Silicon Macs and uses Google's EmbeddingGemma 2 model, according to The Verge. Google says audio, files, and notes never leave the user's computer.
Comments
Loading...