Google Updates Gemini Usage Limits with Compute-Based Quotas
Key Takeaways
- ▸Gemini shifts from simple message counts to compute-based usage limits that refresh every 5 hours with weekly caps
- ▸Usage is now calculated based on prompt complexity, features used (media generation, extended thinking), and chat length
- ▸Premium features like image/video/music generation and extended thinking consume more quota and may cause users to reach limits faster
Source:
Summary
Google has announced changes to Gemini model access and usage limits that introduce a more granular, compute-based quota system. The new system will refresh usage limits every 5 hours until users hit their weekly cap, with calculations factoring in prompt complexity, features used, and chat length—moving beyond simple message counts. Paid subscribers with Google AI plans will receive higher quotas than free users, while premium features like media generation, video generation, and extended thinking will consume more quota. These changes apply to users 18 and older and aim to more accurately reflect the actual computational resources consumed by different types of queries and use cases.
- Paid Google AI subscription users receive higher limits than free users
- Users can manage, upgrade, or cancel their Google AI plans at any time from Gemini apps
