Washington, USA: Google has announced the launch of its new Gemini 4 Carbon AI model, with improvements in programming, complex task execution, and cybersecurity.
Google said the new model topped the Artificial Analysis Intelligence Index in its fifth edition, succeeding the Argon model that the company launched in early October.
The company explained that the model is making a significant leap forward in the field of cybersecurity, thanks to its ability to detect security vulnerabilities in software and write more secure code.
The model topped the SEC-bench Pro test, which measures the ability of artificial intelligence models to detect real errors and vulnerabilities in software, in addition to the ExploitGym test, which measures their ability to transform discovered vulnerabilities into practical means of exploitation.
Gemini 4 Carbon is currently available to a limited group of companies specializing in cybersecurity, with Google later expanding its availability to include additional partners.
