Google is set to restrict users of its free Gemini AI service to the Flash-Lite model from October 9, as the company changes how its artificial intelligence models are accessed across its subscription tiers.
Under the new structure, users without a Google AI subscription will lose direct access to the more capable Gemini Flash and Gemini Pro models. Google AI Plus subscribers will also lose access to Pro, while AI Pro and AI Ultra subscribers will retain access to the more advanced model tiers.
The change marks a significant shift in how Google distributes access to Gemini’s more powerful capabilities. Flash-Lite is designed as an efficient model for speed and everyday tasks such as summarisation and brainstorming, while Flash is built to provide a stronger balance between speed and reasoning. Gemini Pro is designed for more demanding tasks, including complex maths, coding and working with text, files, images and videos.
Changes across Gemini plans
From October 9, free Gemini users will be limited to Flash-Lite. Users on the Google AI Plus plan will continue to have access to Flash but will lose Pro access, with Google expected to notify affected subscribers by email about when the change will apply to their accounts.
AI Pro subscribers will retain access to Gemini Pro, while AI Ultra users will continue to receive the highest levels of access and usage limits.
Google currently says usage limits vary by subscription, with AI Plus offering twice the standard limits, AI Pro four times the standard limits and AI Ultra offering substantially higher limits depending on the subscription.
Impact on complex AI tasks
The changes are likely to affect users who rely on Gemini for advanced reasoning, coding, research and other demanding workflows.
Gemini Pro is specifically designed for complex maths and coding prompts and has a deeper understanding across text, files, images and videos. This means users who have relied on Pro for intensive study, research, software development or document analysis may need to move to a higher subscription tier to retain access to those capabilities.
For free users, the move to Flash-Lite means the model available to them will be optimised primarily for speed and everyday tasks rather than the more demanding workloads associated with Flash and Pro.
Google introduces different thinking levels
Google is also expanding the ability to control how much computational effort Gemini uses when responding to prompts.
Depending on the model and plan, users can access different thinking levels. Standard thinking is designed for most questions, while Extended thinking allows Gemini to spend longer working through complex problems. Deep Think, which requires the Pro model, provides maximum parallel reasoning and is reserved for AI Ultra users according to Google’s current documentation.
Higher thinking levels consume more of a user's available usage allowance, meaning users may need to balance response depth with their subscription limits.
Gemini 4 Argon and future AI access
Google is also positioning its highest subscription tier as the gateway to its most advanced AI capabilities, as the company continues to develop the next generation of Gemini models.
The shift towards subscription-based access highlights Google's broader strategy of differentiating Gemini according to computing requirements, usage limits and the complexity of tasks users want the chatbot to perform.
For millions of people using Gemini without a paid AI plan, however, October 9 will represent a notable change: access to the more capable Flash and Pro models will end, leaving Flash-Lite as the standard model for free users.
No comments:
Post a Comment