New Usage Limits for Google’s Gemini AI Models Announced
In a recent update, Google has announced upcoming changes to the access of AI models within the Gemini application, particularly affecting free users and subscribers to the AI Plus tier. This modification is set to take effect from October 9, and is aimed at redefining usage limits across different subscription plans.
Subscription Plan Changes
As per the latest support documents, users of the Gemini app without a Google AI subscription will now only have access to the Flash-Lite model (version 3.5). Both the well-regarded Flash (version 3.6) and Pro (version 3.1) models will be unavailable to these users moving forward.
AI Plus subscribers, currently billed at $4.99 per month, will also see a change. This group will retain access to the Flash-Lite and Flash models, but will lose access to Pro. Affected subscribers can expect further communication detailing these changes via email.
| Gemini Model | Google AI Plan | Flash-Lite | Flash | Pro |
|---|---|---|---|---|
| Without a plan | ✓ | ✓ | X | |
| AI Plus | ✓ | ✓ | X | |
| AI Pro | ✓ | ✓ | ✓ | |
| AI Ultra | ✓ | ✓ | ✓ |
Pro and Ultra Model Options
Subscribers paying $19.99 per month for the AI Pro and AI Ultra plans will see no decline in the models available. Notably, these users will gain access to the Deep Think feature, which is currently only available on high-tier plans costing $99.99 and $199.99 per month. This feature enhances the AI’s capability for maximum parallel reasoning.
Enhanced Features and Levels
In addition to the changes in model accessibility, the Gemini app is introducing new “effort” levels categorized as low, medium, and high. These new settings, set to roll out in October, will allow users to dictate the complexity of tasks performed by the AI and will, however, draw upon their defined limits.
This will increase the model’s ability to complete tasks and provide more thorough answers to queries, but will also use more of your limit.
Currently, users can enable a feature labeled “Extended thinking: Complex problem solving.” The newly introduced levels will align with settings available in Google AI Studio and Antigravity.
Future Developments: Gemini 4 Argon
As the Gemini platform evolves, the introduction of version 4, named Argon, has also been announced. This will initially be available to AI Ultra subscribers, but details regarding its classification—whether it will join the Pro tier or create a new tier—have yet to be disclosed.
Conclusion
These recent updates signal a significant shift in how Google’s Gemini models will be accessed by various user tiers. With an emphasis on balancing user limits and enhancing capabilities, the forthcoming changes are poised to impact the overall user experience within the app.

FTC: We use income-earning auto affiliate links. More.
