Google Opens Gemini 4 Argon to Paying Developers and Subscribers

Gemini 4 Argon is now open to a much wider audience, after Google first introduced the model on 30 September for trusted cyber defenders only. On 13 October, Google made the model generally available to paid Gemini API developers and to Google AI Pro and Ultra subscribers in the Gemini app and in AI Mode in Search. The model is identified as gemini-4-argon in the API, and it is aimed at complex work across software engineering, enterprise knowledge tasks, and cyber defence. Google describes Argon as its strongest Gemini model yet for following style instructions and handling long, multi-step workflows.

Pricing is a central part of the launch. Gemini 4 Argon costs $2 per million input tokens and $10 per million output tokens through 30 November, with cached input at $0.10 per million tokens. From 1 December, those prices rise to $4 and $20, with cached input at $0.20. There is no free tier for the model on the API, which sets it apart from some other Gemini models. Google lists a context window of 1,048,576 tokens, shared by input and output, and a default maximum output of 65,536 tokens. The model accepts text, images, video, audio, and PDF files as input, and produces text output.

Gemini 4 Argon targets coding and enterprise work

Gemini 4 Argon is positioned for developers and businesses that need a model to work through detailed documents, code, and research tasks without losing track of the goal. Google says the model ranks first on the Vals Index and scores 87 percent on the ERQA embodied reasoning benchmark, figures drawn from Google’s own published testing. The company also reports strong results on safety testing for indirect prompt injection. For subscribers, access depends on the plan: Ultra subscribers get the widest access, including AI Studio and Antigravity, while Pro subscribers can use it in the Gemini app. The wider release marks a shift from a controlled preview to a commercial product, and it gives Google a flagship model to compete with other leading AI systems in coding and business use.

Source: The AI News Report