Google has made significant changes to its AI subscription plans, drawing both praise and criticism from users. Following its I/O 2026 announcements, the company quietly updated the terms for its AI Pro plan, imposing stricter usage limits and introducing a new credit system. At the same time, Google slashed the price of its top-tier AI Ultra plan by $50, while also adding a new $100 per month mid-tier plan. The moves come as competition in the AI space heats up, with providers like Anthropic and OpenAI also adjusting their offerings.

What Changed in the AI Pro Plan?

Until recently, the AI Pro plan offered a relatively straightforward usage allowance, typically measured in tokens or queries per day. However, Google has now transitioned to a credit-based system that calculates consumption based on the complexity of the user's prompt, the features utilized, and the length of the chat session. This means that a simple request might consume fewer credits, while a complex multi-step prompt could drain credits much faster.

According to an email sent to subscribers, the new credit system is designed to be more equitable. Google states that it "more accurately reflects the compute used by each request." However, early adopters on social media platforms like Reddit have expressed frustration. One user reported that a single text prompt consumed 13% of their monthly quota, leading to accusations that the new system is a "scam."

Stricter Time Windows and Weekly Caps

In addition to the credit system, Google is now enforcing usage limits within five-hour rolling windows, a model already used by Anthropic for its Claude plans. After each five-hour cycle, the usage counter resets, but overall usage is capped by a weekly limit. Once the weekly limit is reached, the user's access is restricted until the next cycle begins. These limits apply not only to Gemini but also to other Google AI products such as Antigravity and Flow.

Google has also discontinued the 1,000 free AI credits it previously offered for Flow. Despite this, the company assures users that their experience should not change, as the credits were rarely used by most subscribers. Users can monitor their remaining credits and usage history by navigating to the Usage limit option in Gemini's settings.

Ultra Plan Price Cuts and New Tier

At the I/O 2026 conference, Google announced a price reduction for its AI Ultra plan from $200 per month to $150. Additionally, a new $100 per month Ultra plan was introduced, offering fewer features and lower limits than the full Ultra but more than the Pro plan. These adjustments aim to make advanced AI capabilities more accessible, though some analysts note that the hidden changes to the Pro plan might offset the goodwill generated by the price cuts.

Google has not disclosed the exact number of tokens or credits available under the Pro plan, but it states that users on this tier receive four times the allowance of the free tier. The free tier itself has very limited capabilities and is primarily intended for casual users.

User Backlash and Comparisons to Competitors

The revised AI Pro plan has sparked a wave of complaints, especially from power users who rely on the service for daily tasks. Many argue that the new credit system is opaque and punitive. Unlike the previous model where a user could roughly estimate their daily usage, the complexity-based credits make it difficult to predict costs. Some users have already started exploring alternatives, such as OpenAI's ChatGPT Plus or Anthropic's Claude Pro, which offer more transparent usage policies.

This is not the first time Google has faced backlash over AI pricing. Earlier this year, the company tested weekly limits on a subset of users, a move that was met with similar criticism. Now that those limits are officially implemented, the company appears to be treading carefully, emphasizing the fairness of the new system while monitoring user feedback.

Why Google Made These Changes?

From a business perspective, the changes align with Google's broader strategy to manage compute costs. Large language models require substantial server resources, and as user demand grows, providers must find ways to allocate capacity efficiently. By tying usage to compute intensity, Google can better distribute its infrastructure and prevent a small number of power users from consuming disproportionate resources. The five-hour windows also help smooth out demand peaks.

However, critics point out that the changes also serve to upsell users to more expensive plans. While the $100 Ultra plan is now available, those who need high volumes of complex queries may find themselves forced to upgrade if they exceed Pro limits. The $200 Ultra plan, even with the price cut, remains out of reach for many individual users.

What This Means for Subscribers

Current AI Pro subscribers will automatically transition to the new credit system. Google has indicated that existing users will see no disruption but will need to adapt to the new consumption tracking. The company has not announced any grandfathering of older terms. As a result, users who previously enjoyed unlimited or near-unlimited usage will now have to plan their queries carefully, especially if they rely on AI for tasks such as coding, content generation, or research.

Google's decision to integrate multiple AI services under the same credit umbrella (Gemini, Antigravity, Flow) means that heavy use of one product will affect availability in others. For example, using Flow for data analytics might reduce credits available for Gemini chat sessions. This consolidation could push users to prioritize certain tools over others.

Industry observers note that Google is following a trend set by other AI providers. Anthropic's Claude Pro has long used usage windows, and OpenAI recently introduced more granular limits on its API services. The difference lies in execution; Google's credit system is more flexible in theory but has been marred by poor communication. Many users only discovered the changes through emails or by noticing unexpected slowdowns, leading to frustration.

Historical Context: Google's AI Journey

Google has been a major player in AI since launching its first language models, but it was the introduction of Bard (later renamed Gemini) that brought AI subscriptions to the mainstream. The company initially offered generous free tiers to attract users, then gradually introduced paid plans as demand grew. The Pro plan was launched at $20 per month, offering enhanced capabilities and faster response times. The Ultra plan followed at $200 per month, targeting enterprises and power users.

The I/O 2026 announcements represent a pivot in pricing strategy. By creating a $100 tier, Google aims to capture a segment of users who found the $200 plan too expensive but needed more than the Pro plan provides. The concurrent changes to the Pro plan, however, risk alienating the very users who were the foundation of the paid subscription base. Some analysts predict that Google will need to refine credit calculations to better align with user expectations, or risk losing subscribers to competitors.

Future Outlook: What to Expect

Looking ahead, Google may adjust the credit system based on feedback. The company has a history of iterating on its policies after public outcry. Meanwhile, users are advised to track their usage closely and consider whether their needs might be better met by a higher-tier plan or an alternative service. Google has also stated that usage limits could change without prior notice, so subscribers should stay informed through official channels.

The broader implication is that AI subscriptions are becoming more nuanced. Gone are the days of simple monthly caps. Providers are increasingly tailoring plans to differentiate between casual and power usage. For consumers, this means more choices but also more complexity in decision-making. As Google continues to refine its offerings, it will likely shape the competitive landscape for the rest of the industry.

Source: Android Authority News