Google sets five‑hour reset for Gemini usage limits after I/O announcement

Google changed how Gemini usage limits work on May 20, shifting most users from a daily prompt count to a system that refreshes every five hours. The company’s support documentation says Gemini Apps now use compute-based limits that depend on prompt complexity, the models and features used, and the length of a chat. Google also says those limits continue refreshing every five hours until a user reaches a weekly cap. The change was announced alongside broader Gemini updates at Google I/O 2026 and began showing up in user-facing help pages and dashboards this week. ### What exactly changed in Gemini’s limit system? Google’s Gemini Apps help page says the service no longer centers limits on a flat number of prompts. Instead, the company now measures usage by compute, meaning a short text request and a heavier task such as video generation do not count the same way. The support page says the calculation factors in prompt complexity, features used and chat length, and that the limit refreshes every five hours until a weekly ceiling is reached. (support.google.com) May 20 was the first day many users saw the new framing tied directly to Google I/O announcements. Android Police, citing the I/O rollout, reported that Gemini had moved away from the older daily-prompt model that had distinguished it from rivals including ChatGPT and Claude. ### Does this mean every user gets the same five-hour reset? Google’s support materials say no. (support.google.com) The company says Gemini Apps limits vary by plan, and users can upgrade to paid Google AI plans for broader access to models and features. Google’s consumer help page describes the five-hour refresh as part of a wider quota system rather than a single universal allowance. Google also maintains separate rules for some groups. (androidpolice.com) The reporting cited in the original coverage said under-18 users keep prior rules, unpaid users remain on standard caps, and AI Plus subscribers receive limits that are twice those of free users. Google’s support page confirms that plan-based differences remain in place, though it does not present all tiers in the same wording used by secondary reports. (support.google.com) ### Why are some users saying they hit limits faster? Google’s own explanation points to the new accounting method. Because usage is now tied to compute, a longer conversation or a more demanding feature can consume quota faster than a simple text exchange. That means two users sending the same number of prompts may not see the same remaining allowance. Published reports on May 20 said some subscribers were already running into lockouts within hours of the change. (support.google.com) Those reports described complaints from paying users who said the new system felt stricter than the old prompt-based framing, especially when heavier features were involved. ### Is Google applying this only to Gemini, or more broadly across AI plans? (support.google.com) Google’s May 20 subscription update said its AI plans now come with different usage tiers, including higher limits for top-end subscribers. In that post, Google said its Ultra plan offers a usage limit five times higher than Pro in the Gemini app and related tools, and that the company also cut the monthly price of AI Ultra to $200 from $250. (msn.com) Google’s I/O roundup points users to the company’s Gemini announcements and product pages for the rest of the rollout. The clearest next place to watch is the Gemini Apps help center, where Google is updating plan descriptions, quota language and feature availability as the new system rolls out. (blog.google 1) (blog.google 2)