Google released Gemini 4 Carbon on Wednesday, a new AI model that the company says offers improvements in coding, agentic tasks and cybersecurity compared to its earlier Gemini 4 Argon model released earlier this month, according to The Verge.
The model is currently available only to a select group of cybersecurity defense companies, with plans to expand to other partners in Google’s Fairwind Program. Google claims Carbon ranks first on the Artificial Analysis Intelligence Index (v5.0) benchmark, ahead of Claude Opus 5.5 and GPT-6 Astra.
Google says Carbon holds the top position on SEC-bench Pro, which measures AI model ability to discover real-world bugs, and ExploitGym, which measures how models turn vulnerabilities into exploits. The company positions the model as better at finding vulnerabilities and generating secure code.
According to Google, Carbon is designed for longer tasks, with internal teams seeing it “run for days to persist through dead ends to tackle complex, multi-step engineering tasks.” The company also claims the model is better at avoiding shortcuts, resisting prompt injection attacks and collaborating with users.
Google says Carbon offers more “transparency and honesty” by cutting “unsupported claims” by two times when compared to Argon. The model will be priced at $2 per million input tokens and $10 per million output tokens, matching Argon’s introductory pricing.