Gemini User Burns 5-Hour Cap in One Prompt
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google rolled out new compute-based usage limits for Gemini, replacing fixed prompt caps with a credit-style system that weighs prompt complexity, features used, and conversation length, refreshing every five hours under the AI Pro plan.
- Ashutosh Shrivastava posted video proof on X showing his five-hour rate limit jumped from 0% to 100% after one avatar-based video-generation prompt that ran three to four minutes and ultimately failed to produce a video.
- Josh Woodward, Google's Gemini lead, publicly responded to the complaint with "Yikes, let us take a look!" but had not announced a policy change at the time of writing.
- Reddit users on the Gemini subreddit have piled on with similar complaints, saying the new quotas feel significantly tighter than the previous fixed-prompt system and make it hard to predict how much usage a single task will consume.
- Google has boosted Gemini quotas for Antigravity users by up to 9x compared to the immediate post-nerf period, but broader limits for regular AI Pro subscribers remain unchanged.
Why it matters: Paying AI Pro subscribers are losing trust in Gemini's reliability when a single failed prompt can drain a five-hour quota — a concrete breakdown of the value proposition Google sells with its subscription tier. Google's selective workaround for Antigravity users (9x higher limits) while leaving regular subscribers on unchanged caps signals the company knows the system is too tight but hasn't committed to a broader fix.
Ask SkimNews
