According to a report by tech journalist David Nield in WIRED magazine, earlier this summer Google rolled out a series of significant upgrades to its Gemini artificial intelligence applications, making them more powerful and far-reaching than ever before. Google’s AI is now capable of working across a broader range of applications, and using Google's various products is becoming something that is very difficult to do without encountering some kind of AI feature or assistant of its making. However, alongside these AI system upgrades, new usage limits have also come into effect. Google has completely changed the way AI usage is measured and calculated across its different subscription tiers: the Free tier, the Plus tier (AI Plus), the Pro tier (AI Pro), and the Ultra tier (AI Ultra). If you have suddenly found yourself in a situation where you have run out of "credit" in the Gemini AI bank and been asked to wait before making further prompts, this is likely the cause. Below is a detailed explanation of the changes and how you can check where your usage balance stands.
How AI Usage Is Changing
Google now measures Gemini AI usage based on the computing power requirements of your prompts, rather than the raw number of requests you make. Consequently, while in the past you might have been able to generate three videos a day, you may now find that you can only generate two if they are particularly complex. From Google’s perspective, this approach makes more sense because it accurately measures how much you are actually costing the company and its data centers in terms of resources and energy consumption.
However, for end users, the new method remains somewhat vague and makes it difficult to know exactly when they are likely to hit their limits. The practical meaning is that users can no longer rely on a simple, fixed rule of thumb like "I can perform 5 image generations a day." Furthermore, Google explicitly notes in its official support documents that access to services is subject to change or may be limited based on testing, experimentation, or system availability. In simple terms, this means that the level of activity available to users may change significantly from day to day depending on testing status and system load.
The two main factors that determine how much AI you can consume are the subscription tier you are on, alongside the complexity and length of your prompts. For example, a simple request for a weather forecast will require very few resources, whereas a request to write code for a small app will consume significantly more resources. In addition, the specific Gemini AI model you choose to use (for example, Gemini 3.5 Flash) also affects resource consumption. Users can select their preferred AI model directly from the prompt input box in the application.
The Quotas for Each Plan
For users located in the United States, Google offers four Gemini AI subscription plans to choose from, or alternatively, the option to continue using the free tier. The paid subscription plans include the AI Plus plan at $8 a month, the AI Pro plan at $20 a month, and the AI Ultra plan at either $100 or $200 a month. The higher your monthly payment, the larger the volume of AI usage allocated to you, allowing you to use more advanced AI models for more extended periods.
Google does not specify in its official documents what the usage limits are on the free tier, describing them solely as "standard" limits. Beyond that, users on the AI Plus plan receive a quota that is 2x these standard limits, while AI Pro plan users enjoy a quota 4x the standard limit. The limits for AI Ultra subscribers are either 5x or 20x higher than those of the AI Pro plan, depending on the specific payment level the user is on (a $100 subscription or a $200 subscription).
It is worth noting that all users have access to the full range of Gemini AI models, including the Flash-Lite, Flash, and Pro models. As you progress up this chain of models to those with more advanced and smarter capabilities, they consume more computing power and count more heavily against your remaining quota. Each model also features different "thinking" levels—Standard, Extended, and Deep Think—which directly affect response quality, response speed, and practical usage limits.
Context Windows of the Different Models
The final significant difference between the various models and subscription plans concerns the size of the context window. Essentially, this figure indicates and measures exactly how much information you can include in a single, continuous conversation thread with the artificial intelligence. For users on the free tier, the context window stands at 32,000 tokens (tokens are small pieces of text), equivalent to approximately 24,000 words in total.
For users registered on the AI Plus plan, the context window jumps to 128,000 tokens, which is roughly 96,000 words. For users on the most advanced and expensive plans, AI Pro and AI Ultra, the limit stands at a full 1 million tokens—equivalent to approximately 750,000 words in a single conversation—a figure that allows inputting particularly long files and documents into the conversation interface without losing the conversation's context.
How to Track Actual Quota Utilization
While the new rules Google has set around AI usage may lack precise numerical details for some of the plans, in practice, it is very easy to check where you stand at any given moment. In the Gemini app running on a web browser (the web version), you need to click on the cog icon located in the lower-left corner and then select the "Usage limits" option. In the mobile app designed for Android or iOS operating systems, you should first tap the menu button located in the top-left corner, go from there to the cog icon, and then tap the "Usage limits" option.
Upon entering this screen, you will be presented with two clear bars detailing your current status:
- The top bar shows your current usage at that moment, which resets automatically every five hours. If you have utilized the full quota of these five hours, you will need to wait before making further prompts, and the Gemini app will display the exact next reset time on your screen.
- The second bar displays your weekly usage limit, which resets every week (and is also clearly displayed on the screen).
If you subscribe to a paid plan and reach these maximum limits, the system will not block you completely but will instead demote you to using Google’s most basic AI model, which you can continue to use as normal until the next reset time arrives. Naturally, on the usage limits screen, you will regularly see offers and invitations to upgrade your AI plan to a more expensive one. Furthermore, it is important to take into account that Google’s official support documents emphasize that these limits may change entirely without prior notice due to capacity constraints and loads in the data centers, and that users on the free tier are expected to be the first to be affected by these changes if Google needs to manage and reallocate its AI resources due to high demand.