Google has released Gemini 4 Carbon, a frontier model designed specifically for cybersecurity defence companies. The announcement on Wednesday highlights significant gains in coding and agentic tasks, with the model currently topping the Artificial Analysis Intelligence Index v5.0 benchmark ahead of competitors like Claude Opus 5.5 and GPT-6 Astra. This new version improves upon the previous Argon release, though it remains restricted to a select group of partners within the Fairwind Program before wider availability.
The release matters because it signals a shift in how large language models are evaluated and deployed within high-stakes security environments. By prioritising performance on the Intelligence Index, Google emphasises that specific benchmarking matters more than general capability claims for this sector. The model is not yet open to the public or general enterprise users, meaning its practical impact on broader API behaviour or customer privacy remains unproven at this stage.
* Available exclusively to select cybersecurity partners via the Fairwind Program
* Currently ranks first on the Artificial Analysis Intelligence Index v5.0
* Represents an upgrade over the earlier Argon model released this month
