Washington, November 01, 2026 : Google has unveiled Gemini 4 Carbon, its latest artificial intelligence model, featuring advancements in software development, complex task handling and cybersecurity.
According to the company, the model secured the top position in the fifth edition of the Artificial Analysis Intelligence Index, replacing Google's Argon model, which was introduced earlier in October.
Google highlighted the model's enhanced cybersecurity capabilities, particularly its ability to identify software vulnerabilities and generate more secure code.
Gemini 4 Carbon also achieved leading results in the SEC-bench Pro benchmark, which evaluates AI models' ability to detect real-world software errors and security weaknesses. It also topped the ExploitGym assessment, which measures how effectively models can turn identified vulnerabilities into practical exploits.
The model is currently available to a select group of companies specialising in cybersecurity, with Google planning to extend access to additional partners at a later stage.
Source: QNA
