Google Gemini’s switch to compute billing sparks user uproar after quota-draining glitch

Billing change and instant backlash

Google has moved Gemini from a simple “prompts-per-day” quota to a compute-based billing model, and users are angry. It has been reported that Android Authority and Chinese tech outlet IT Home (IT之家) flagged a wave of complaints after Google adjusted Gemini’s quota algorithm — reportedly shifting charges to reflect compute consumption, request complexity, invoked functions and even chat-history length. The company also ties these calculations to its Google AI Pro tier, a $20-per-month subscription, amplifying concerns about opaque metering.

The incident that lit the fuse

What set off the firestorm? It has been reported that a user, Ashutosh Shrivastava, posted a video showing that a single, simple prompt to generate a headset video consumed an entire five-hour quota in about 3–4 minutes, despite the account showing 0% usage beforehand. The video reportedly did not produce a finished result. Gemini lead Josh Woodward publicly shared the clip and said the team would investigate. The episode has become the focal point for wider anger over unpredictable charges and unclear limits.

Why this matters beyond one glitch

For Western readers less familiar with how China follows global AI moves: major Chinese platforms and enterprises closely watch Gemini as part of a broader contest between US and Chinese AI stacks — and they track monetization models as much as model quality. Google’s services are broadly inaccessible inside mainland China, yet Chinese media and developers monitor these changes for clues about pricing pressure, cloud costs and industry direction. With rising compute demands and growing regulatory scrutiny on both sides of the Pacific, will users accept less transparent metering? For now, Google says it will probe the incident while criticism continues to mount.