Google opens Gemini 4 Argon to all developers

Google made its new AI model, Gemini 4 Argon, available to developers around the world on October 13, 2026. Developers can use it through the paid tier of the Gemini API, in AI Studio and Android Studio. Businesses can use it through Google's enterprise products. Google first showed the model on September 30 to a small group of cybersecurity partners.

According to Google, the model can handle about one million tokens at once. Google's developer page gives the exact limit as 1,048,576 tokens, shared between input and output. By default, a single answer is limited to 65,536 tokens, but a "long decode" option allows much longer outputs. HPCwire notes that rival models usually stop at around 128,000 output tokens per request.

The starting price is 2 dollars per million input tokens and 10 dollars per million output tokens. Google says the price will rise to 4 and 20 dollars from November 30, 2026. Paying subscribers of Google AI Pro and Ultra can also use the model in the Gemini app and in AI Mode in Search.

Google says Argon ranks first on the Vals Index, which tests work in finance, law, tax and coding. It also reports strong results in robotics reasoning and in resisting hidden instructions, known as prompt injection. These scores come from Google. HPCwire and other outlets did not report independent tests yet.

Independent researchers have not yet published their own comparisons, so it is not yet known how the model performs outside Google's tests.