Google Shifts Gemini to Compute-Based Usage Limits, Raising Cost of Premium Features
Key Takeaways
- ▸Gemini usage now resets every 5 hours on a compute-based model rather than simple request counts
- ▸Google AI premium subscribers receive higher usage limits than free-tier users
- ▸Premium features (media generation, extended thinking, Deep Think) consume more usage credits, incentivizing paid upgrades
Source:
Summary
Google has announced significant changes to Gemini's usage model, transitioning from traditional metrics to compute-based limits that reset every 5 hours. The new system calculates usage based on prompt complexity, feature use, and chat conversation length, creating a more nuanced allocation system. Paid Google AI subscription users will receive higher limits than free users, and premium features—including media generation (images, videos, music), extended thinking, and Deep Think—will consume usage allowances faster than standard interactions. The changes apply to users 18 and older and reflect Google's strategy to balance service accessibility with resource constraints as AI demand accelerates.
- Usage calculations factor in prompt complexity, feature type, and conversation length
- Changes apply only to users 18 and older
