Gemini 4 Argon Writing Code Instead of Programmers? / See Google's Striking Figures

Using Google's new model in coding tools increased shipped code by over 35 percent and reduced rollbacks to the previous version by 30 percent.

Google has unveiled the Jumnai 4 Argon artificial intelligence model; A model designed for developers and businesses, targeting the ability to perform complex and multi-step tasks. This model has been released in a situation where the competition between big technology companies to develop advanced artificial intelligence tools and attract corporate customers has intensified. Jumnai 4 Argon has been developed for applications such as programming, finance, law, science and cyber security.

Google's focus is that artificial intelligence can perform a series of related tasks beyond answering questions and help users in complex work processes. According to Google, the use of this model in the antigravity programming tool has increased the sending of codes produced with the help of artificial intelligence agents by more than 35%. Also, returning to previous versions of the code in this process has been reduced by almost a third. These statistics show that in addition to the capabilities of the model, Google is also focusing on improving the performance of tools based on artificial intelligence agents. In terms of pricing, Google has set the cost of using the Jamna 4 Argon programming interface at $2 for every one million input tokens and $10 for every one million output tokens.

These tariffs are going to be doubled after November 30; An issue that can affect the cost of using this model for organizations and developers. Companies such as Deloitte and LexisNexis have also started testing this model.

The presence of these companies shows that Google seeks to expand the use of artificial intelligence in professional and organizational environments; A market where companies such as OpenAI and Anthropic are also competing to acquire a larger share of it. The supply of Jamna 4 Argon is part of Google's effort to strengthen its position in the artificial intelligence market. The success of this model will ultimately depend on its performance in real-world tasks, developer acceptance, and its ability to deliver economic value to businesses.