Google recently changed the way it controls the use of Gemini. Instead of maintaining a fixed number of requests, the platform now considers the total processing expense and resets the limit every 5 hours. The change adopts a model that charges for the computational cost, but has generated significant complaints among users who report excessive consumption of daily allowances.
A user recorded on X that he reached the 5 hour limit in a few minutes with just a simple command. The post caught the attention of Josh Woodward, leader of Google Labs and the Gemini application, who responded expressing surprise and stated that the team would investigate the case. The situation highlights problems in the implementation of the new access control system.
Consumo of Tokens Challenges Users on the Platform
Prompts complexes now face direct impact under the new model. Usuários report that tasks that previously consumed few tokens now quickly exhaust available allowances. Mensagens on Threads and X document growing complaints about the change, with many pointing out that basic operations are unexpectedly pushing boundaries.
Google identified high token consumption when using seemingly simple prompts as one of the main issues. Para try to resolve the issue, the company released the variant Gemini 3.5 Flash (Low) on the Antigravity platform, a tool used as a programming assistant. In internal testing, the Low version recorded up to 45% lower token consumption compared to the standard variant, offering a more economical alternative for users concerned about daily allowances.
Modelos alternatives and optimized version reduce cost
The release of Gemini 3.5 Flash (Low) represents an attempt by Google to balance functionality and resource consumption. Usuários who need to perform repetitive or less demanding tasks can now use this variant without compromising their daily allowances. The 45% reduction in token consumption makes the new version attractive to programmers and developers who frequently use code assistance.
Além Additionally, Google reinforces that certain advanced features accelerate limit spending. Media (images, files, and music) Geração, Deep Research, and other advanced modes of Gemini consume significantly more tokens than standard operations. Usuários who exploit these features should expect faster reduction of their available daily allowances.
- Gemini 3.5 Flash (Low): 45% lower consumption than the standard variant
- Recursos that accelerate consumption: imaging, Deep Research, advanced modes
- Limits Resetamento: every 5 hours instead of daily
- Plano AI Ultra: offers a limit up to 20 times higher than the standard version
Why Google changed the billing model
Ainda there is no official explanation from the company about the reasons for the change. Analistas indicate that the market will likely shift to fee-for-process models to adapt to the era of AI agents. Quando an agent can execute long actions in the background, it would not be viable to charge for each attempted use, but rather for the processing spent on the task.
Esse tuning represents a trend in the artificial intelligence sector, where companies like OpenAI, Anthropic and Google test different monetization strategies. Cobrar for real processing offers greater transparency and aligns costs with the value delivered to the user. However, the transition created confusion because the system does not clearly communicate how much processing each action consumes.
Premium Planos offers greater security for users
Google offers the Ultra AI plan as an alternative for users who face frequent limitations. Esse plan can offer a limit up to 20 times higher than the standard version, ensuring continuous use without interruptions due to excessive consumption. Usuários for professional applications or intensive development can justify the cost of the upgrade to avoid service outages.
The plan scaling strategy allows Google to maintain a viable free tier while monetizing users with higher needs. Quem depends on Gemini for daily work may consider investing in paid plans a practical solution. The price and specific benefits of the Ultra AI plan were not detailed in the company’s recent communication, leaving users without complete information for purchasing decisions.