Skip to content
IFAI

AI tools

Gemini Notebook ditches daily limits for 5-hour refreshes

Published Reach 37

Google just introduced flexible, compute-based usage limits for Gemini Notebook, replacing the rigid daily reset with refreshes every five hours. The system factors in prompt complexity, chat length, and feature usage, and lets you queue heavy tasks like slide decks when running low. It rolls out to consumer accounts on web and mobile starting Sept 2.

Gemini Notebook ditches daily limits for 5-hour refreshes
Image: blog.google

If you rely on Gemini Notebook to tear through heavy research, hitting your daily ceiling before lunch is a familiar pain. Google is finally rethinking how that limit works.

Dynamic compute budgets and 5-hour resets

Starting September 2, Google is rolling out compute-specific, flexible usage limits to Gemini Notebook for consumer accounts across web and mobile.

Instead of resetting once a day, usage limits will now refresh every five hours. That single change makes it much easier to keep working across an entire day without grinding to a halt.

Your allowance is now metered dynamically. Google factors in prompt complexity, conversation length, the number of referenced sources, and which features you invoke.

If a request is too heavy for your remaining limit, the notebook will suggest lighter output alternatives. When you run out of juice generating intensive assets like Video Overviews or Slide Decks, you can defer them. The system will queue them up, run them automatically once your quota resets, and send a notification when they are ready.

Why it matters

Moving to a five-hour refresh cycle solves the most frustrating roadblock for regular users, effectively killing off the dreaded dead zone after a busy morning session. Deferring heavy generative tasks with background notifications is also a smart piece of product design that treats compute like an asynchronous queue instead of a brick wall.

That said, compute-based metering cuts both ways. Power users who regularly feed dozens of long documents into a single prompt will likely burn through their budget much faster than before. You will need to be a bit more deliberate about how much context you throw at the model at once.

Sources (1)

  1. [1]
    blog.google(opens in a new tab)Originalblog.google

Best of AI, daily