Google today revealed its next frontier AI model, which it calls Gemini 4 Argon. The new model offers “state-of-the-art performance in complex workflows in real-world software engineering, business knowledge work such as legal and financial, and cybersecurity defense,” according to chief AI architect and Google DeepMind senior vice president Koray Kavukcuoglu. But the company is limiting access at first to a “set of trusted cyber defenders,” and Kavukcuoglu says Google is “actively participating in the US government’s voluntary process for access to the pre-launch model as we gradually expand access.”
Kavukcuoglu says Gemini 4 Argon is already powering Google’s “internal workflows,” highlighting how the model has helped with tasks like large-scale codebase migrations. Google also shared a large graph showing many benchmarks where the Gemini 4 performs better than competing models from OpenAI and Anthropic. (The AI community was abuzz earlier this week with X over what appeared to be leaked benchmarks for the new model.) Google will strengthen “critical border safeguards” ahead of a broader rollout, including defending against misuse and fast injection attacks and monitoring for misalignments.
