Google Unveils Gemini 4 Argon AI Model with Frontier-Level Capabilities

Google has unveiled its latest AI model, Gemini 4 Argon, which it claims outperforms rival models in coding, knowledge work, and cybersecurity benchmarks. The model is already being used by Google for safer code migration and complex quantum computing research tasks.

The Vals Index vibecoding benchmark shows that Argon can create applications from scratch using human language, and the model is leading in this area according to Google's metrics.

Argon is also capable of finding, verifying, and patching critical software flaws without human input, which could potentially reduce the risk of AI models being exploited.