Google announces Gemini 4 Carbon, its latest frontier AI model that you can't use

Google has announced Gemini 4 Carbon, a new frontier AI model that builds on the capabilities of Gemini 4 Argon. Carbon improves upon Argon in coding, agentic tasks, and scientific analysis, and is particularly strong in cybersecurity applications, showing a 'significant leap' in vulnerability detection, validation, and remediation. Carbon is also twice as likely to generate code free of vulnerabilities compared to Argon, and is 10 times better at generating 'high-quality' patches than non-Google models. Carbon has achieved a score of 89 points on WebDev Arena, ranking it at number four on the leaderboard, and beats several other models, including GPT-6 Astra, Claude Opus 5.5, and Claude Fable 5.1, in almost every test. However, Google is not making Carbon available for public use due to its capabilities.