New Delhi, October 5: Google is preparing to restrict access to some of its AI models for consumer users, with tighter limits planned for both free accounts and entry-level AI Plus subscribers. According to an updated support document, the changes build on Google’s move towards compute-based usage tracking introduced in May.
From October 9, users accessing Gemini on the web and mobile apps without an active Google AI subscription will only be able to use the 3.5 Flash-Lite model. They will lose access to the faster 3.6 Flash and higher-capability 3.1 Pro models. AI Plus subscribers will retain access to Flash-Lite and Flash, but will no longer be able to use the 3.1 Pro model.
There will be no model removals for AI Pro subscribers. The tier will also gain access to Google’s advanced Deep Think feature, which offers maximum parallel reasoning and was previously limited to the higher-end $99.99 and $199.99 monthly plans. AI Ultra subscribers will continue to have access to Deep Think and early access to upcoming model rollouts, with no reductions in their existing capabilities.
Google is also introducing usage limits for media generation, including images, videos and music, as well as Deep Research and the Pro model. “Gemini will move to compute-based usage limits that will refresh every 5 hours until you reach your weekly limit,” Google said. The company added that usage will depend on factors including prompt complexity, features used and chat length.
Paid users will have higher usage limits than those without a Google AI subscription, while premium models and features will consume more usage and could cause users to reach their limits faster.
