Google Launches Gemini 4 Argon AI Model for Professionals
Google released Gemini 4 Argon, a high-performance AI model for complex professional workflows, featuring competitive introductory API pricing to rival OpenAI and Meta.
Google released its flagship AI model, Gemini 4 Argon, on October 13, 2026, targeting developers, businesses, and paying subscribers. Positioned as a powerhouse for complex workflows in coding, cybersecurity, finance, law, and science, the model is available to Gemini Pro and Ultra subscribers for at least $19.99 per month and through the Antigravity development platform.
To attract developers, Google introduced introductory API pricing of $2 per million input tokens and $10 per million output tokens, though these rates will double after November 30. Internally, the company reports that Argon increased AI-generated code production by 35% and improved the performance of Chrome and YouTube.
The launch follows a period of internal instability, including leadership changes at Google DeepMind and the cancellation of Gemini 3.5 Pro. The release comes amid intense competition from OpenAI's GPT-6 and Meta's Muse Wonder models. To ensure safety, Google tested the model with trusted cyber defenders and implemented a chain-of-thought tracking system to prevent harmful behavior.