Google launched its latest frontier model, Gemini 4 Argon, to eligible users on Tuesday.
"Argon excels at deep work in complex, real-world tasks across domains including coding, finance, legal, cybersecurity, and science," Google said in a blog post.
Argon has outperformed other frontier models on benchmarks across domains, owing to its strengths in multi-step tasks. In particular, Argon has ranked first on the Vals Index, which measures model performance across critical economic sectors such as finance, coding, legal, and tax work.
Google previously announced on Sept. 30 that it expanded the model's output token limit to 1 million tokens, up from 64,000, to support Argon's capabilities across longer, more complex use cases.
Teams inside Google have already been using Argon and are seeing "significant impact" across solving major engineering problems, optimizing business workflows, discovering and fixing critical cybersecurity vulnerabilities, and enabling novel breakthroughs in quantum computing, nuclear fusion, and life sciences.
According to the blog post, developers and enterprises can now access the model through the API in AI Studio and Gemini. Google AI Ultra subscribers can use the model via Google Antigravity, while Google AI Ultra and Pro subscribers can access it through the Gemini app and AI Mode in search.
The latest model has an introductory price of $2 per million input tokens and $10 per million output tokens, with cached input tokens at 95% off. After Nov. 30, this rises to $4 and $20, respectively.