Google updated the usage limits for its Gemini AI platform in May 2026 following a wave of user complaints [1].
These adjustments reflect the ongoing tension between maintaining sustainable compute costs and meeting the demands of a growing user base. As AI tools become more integrated into professional workflows, the stability and availability of high-compute features directly impact productivity.
In response to feedback regarding restrictive quotas, Google increased the compute-based usage limits for its Antigravity tools three times [2]. This move aims to alleviate frustrations from power users who found previous caps too limiting for complex tasks.
However, the company is simultaneously experimenting with different access tiers. Google is currently testing weekly caps that may tighten quotas for users on the free version of the service [3]. This suggests a strategy of prioritizing resources for paying subscribers while limiting the ceiling for non-paying accounts.
To bridge the gap between different service levels, Google introduced the AI Ultra Lite subscription tier [4]. This new plan is priced between the existing $20 Pro plan and the $250 Ultra plan [4]. By creating a mid-tier option, the company provides a path for users who need more than the free tier, but cannot justify the cost of the top-tier Ultra subscription.
These changes rolled out in early May 2026 [5]. While the increase in Antigravity limits provides immediate relief for some, the testing of tighter free-tier caps indicates that the company is still refining its balance of accessibility and infrastructure costs.
“Google increased the compute-based usage limits for its Antigravity tools three times”
Google is shifting Gemini from a broad-access model toward a more rigid tiered monetization strategy. By increasing limits for high-end tools while restricting free access, the company is incentivizing users to migrate toward paid subscriptions. The introduction of the 'Ultra Lite' tier suggests Google is attempting to capture a broader segment of the market by offering a more granular pricing structure to manage the high cost of AI compute.
