Google has unveiled Gemini 4 Argon, its new flagship AI model, designed to handle complex, long-running tasks in software engineering, finance, legal work and cybersecurity.
The model was announced on September 30 and is initially being provided to trusted cybersecurity defenders as Google strengthens its safety measures before a wider rollout.
Google says Argon can sustain reasoning across tasks requiring up to 1 million tokens, significantly above the previous 64,000-token limit. The company also says the model can autonomously identify, validate and patch software vulnerabilities.
The company is initially limiting access because of the model's powerful cybersecurity capabilities and potential misuse. Google plans to expand access to paid API customers and Google AI Ultra subscribers before broader availability.
-AH