Google launches Gemini 4 Carbon, an advanced AI model prioritising cybersecurity and outperforming its predecessor in code security
This model follows the recent release of Gemini 4 Argon as part of Google’s expanding AI offerings for developers and businesses.
Gemini 4 Carbon provides superior performance in cybersecurity tasks, including vulnerability identification and software patching.
New AI model
Carbon achieves over double the success rate of Gemini 4 Argon in generating secure code according to internal benchmarks.
It is tailored for complex engineering and coding challenges.
The model is initially available to selected cybersecurity defenders via Google’s Fairwind Program as further safeguards are tested.
Pricing for Gemini 4 Carbon is set at $2 per million input tokens and $10 per million output tokens, consistent with Gemini 4 Argon’s prices.
Reports indicate that the release reinforces Google’s commitment to innovation in AI and cybersecurity sectors.
The development highlights the ongoing demand for advanced security solutions in the digital landscape.
Google’s strategic moves in AI may significantly shape industry standards and practices moving forward.
