Google restricts Gemini 4 Argon rollout to cyber partners
Google says its new model can find and patch software vulnerabilities, but initial access is limited to selected partners and its performance claims have not been independently assessed in the supplied reporting.
Alphabet has launched Gemini 4 Argon, a model for coding, research and writing, with an initial rollout to selected cyber partners through its Fairwind Program.
Google says Argon was trained for defensive cybersecurity and can autonomously find, validate and patch critical software vulnerabilities. The company has not specified the criteria for selecting partners or when wider access might follow.
Google also describes the model as capable in software engineering and visual analysis. It says staff use Argon for debugging and codebase migrations, and that it can analyse material such as long videos and charts.
The company claims Argon outperformed OpenAI’s GPT-6 Astra and Anthropic’s Fable and Opus on several benchmarks. It also cites Vals’ model index, which ranks Argon first. The supplied reporting provides no benchmark results or independent assessment to verify those claims.
The release comes as major AI developers promote new models as more capable than their predecessors. Google said in August that its Gemini app had more than one billion monthly users.
