Google Introduces Compute-Based Usage Limits for Gemini AI Service

Google has updated its Gemini AI service by introducing compute-based usage limits, effective May 17, 2026. This new system calculates usage based on prompt complexity, features utilized, and chat length, replacing the previous daily prompt limits. Usage limits now refresh every five hours, with an overarching weekly cap. Google states this approach better allocates resources, as simple text prompts consume less compute power than complex video or coding prompts.

The free tier remains available, but paid subscriptions have been restructured. The $7.99/month AI Plus plan offers double the usage limits of the free tier, while the $19.99/month AI Pro plan provides four times the free tier's limits. Higher tiers, AI Ultra at $100/month and $200/month, offer five and twenty times the Pro plan's limits, respectively. Previously, Google did not specify exact usage limits for its AI plans, using terms like "more" and "higher."

Users have expressed dissatisfaction with these changes, particularly regarding the value proposition of the AI Pro plan compared to AI Plus. Some users note that AI Pro offers less usage per dollar than AI Plus, though additional perks of AI Pro should be considered. Others are concerned about the stricter usage limits, despite this being a trend in the industry.

To help users monitor their usage, Google has introduced a dashboard displaying the remaining AI usage. This feature is gradually rolling out globally, allowing users to plan their usage accordingly.

These changes come amid increasing demand for AI services and the associated strain on resources. As AI services expand, the hardware supply shortages have further impacted the availability of resources, prompting companies like Google to implement usage limits.