Google launches Gemini 4 Carbon with a ‘significant leap’ in cybersecurity
Also covered by 14 sources
What happened
Google has introduced Gemini 4 Carbon, a new AI model designed for cybersecurity and advanced coding tasks. The model was first made available to a select group of cybersecurity professionals as part of a staged rollout. Gemini 4 Carbon achieved top scores on multiple industry benchmarks for vulnerability detection and secure code generation. Google plans to expand access to more users after further testing and evaluation.
Key details
- Gemini 4 Carbon features a 1 million token context window and supports text, image, video, and speech inputs.
- The model scored 88.2% on SEC-bench Pro for vulnerability finding and 71.0% on CWE-bench for patching.
- It is currently accessible through the Fairwind Program to trusted cyber defenders only.
- Pricing is listed at $2 per million input tokens and $10 per million output tokens, with launch discounts available.
- Gemini 4 Carbon leads the Artificial Analysis Intelligence Index v5.0 with a score of 59.
Why it matters
Cybersecurity professionals and organizations working to defend against digital threats now have access to a model that outperforms previous AI systems in detecting vulnerabilities and generating secure code. This staged release means that only vetted cyber defenders can use Gemini 4 Carbon for now, potentially improving their ability to respond to cyber risks. Broader access and pricing details will affect when and how other developers and companies can benefit from these new capabilities.