Google Introduces Gemini 4 Carbon with Advanced AI Capabilities
(MENAFN) Google has introduced Gemini 4 Carbon, its newest artificial intelligence model, on Wednesday, highlighting claimed improvements in software development, autonomous agent operations and cybersecurity.
The launch comes only two weeks after the release of Gemini 4 Argon. Google described Carbon as its most capable frontier model to date, emphasizing advanced cybersecurity performance and its ability to handle complex tasks over extended periods, according to reports.
The company attributed the accelerated development schedule to closer collaboration between its research teams and AI systems, saying they were "working hand in hand to advance capabilities faster."
Google also claimed that Gemini 4 Carbon secured the top position in the latest version 5.0 of the Artificial Analysis Intelligence Index, achieving a score of 59.
Drawing on feedback from users of its previous model, the company said it had refined Carbon's performance in full-stack web development, scientific problem-solving and multimodal comprehension. Google also reported that the model climbed to fourth place on the WebDev Arena leaderboard after gaining 89 points.
The company highlighted results across several cybersecurity assessments, reporting an 88.2% score on SEC-bench Pro for vulnerability detection, 54.8% on ExploitGym for exploitation tasks and 71% on CWE-bench for security patching. Google characterized the model's overall performance in these evaluations as a "firm lead."
"On internal benchmarks, Carbon drives a strong uplift in generating secure code, and in live coding environments with our own teams it shows reduced misalignment and fabrication rates," the company said.
The launch comes only two weeks after the release of Gemini 4 Argon. Google described Carbon as its most capable frontier model to date, emphasizing advanced cybersecurity performance and its ability to handle complex tasks over extended periods, according to reports.
The company attributed the accelerated development schedule to closer collaboration between its research teams and AI systems, saying they were "working hand in hand to advance capabilities faster."
Google also claimed that Gemini 4 Carbon secured the top position in the latest version 5.0 of the Artificial Analysis Intelligence Index, achieving a score of 59.
Drawing on feedback from users of its previous model, the company said it had refined Carbon's performance in full-stack web development, scientific problem-solving and multimodal comprehension. Google also reported that the model climbed to fourth place on the WebDev Arena leaderboard after gaining 89 points.
The company highlighted results across several cybersecurity assessments, reporting an 88.2% score on SEC-bench Pro for vulnerability detection, 54.8% on ExploitGym for exploitation tasks and 71% on CWE-bench for security patching. Google characterized the model's overall performance in these evaluations as a "firm lead."
"On internal benchmarks, Carbon drives a strong uplift in generating secure code, and in live coding environments with our own teams it shows reduced misalignment and fabrication rates," the company said.
Legal Disclaimer:
MENAFN provides the
information “as is” without warranty of any kind. We do not accept any
responsibility or liability for the accuracy, content, images, videos,
licenses, completeness, legality, or reliability of the information
contained in this article. If you have any complaints or copyright issues
related to this article, kindly contact the provider above.

Comments
No comment