Google rolls out Gemini 4 Argon globally via the Gemini API and to Google AI Pro subscribers in the Gemini app and AI Mode, and in Antigravity for Ultra users

Google's new Gemini 4 Argon model is available at a substantially lower price point than its main competitors.

Google has launched its Gemini 4 Argon AI model globally, aiming to compete with leading AI systems. The company claims Argon offers enhanced reasoning and performance for complex tasks, while reducing common AI chatbot issues like hedging and repetition. Google states Argon outperforms models like OpenAI's GPT-6 Astra and Anthropic's Claude Fable 5.1 and Opus 5.5 across benchmarks in areas such as coding, cybersecurity, science, finance, and law. Internally, Google is using Argon for tasks like identifying cybersecurity vulnerabilities, optimizing YouTube's comment processing, and debugging for its drone delivery company, Wing. The model is now available to Google AI Pro and Ultra subscribers via the Gemini app, AI Mode in Search, and the Antigravity platform. Google is offering an introductory price of $2 per million input tokens and $10 per million output tokens, significantly undercutting its competitors, with these prices expected to double after November but remain competitive. This release marks Google's first major frontier model update since Gemini 3 in late 2025, following a period where rivals released notable models.

Google's release of Gemini 4 Argon re-establishes its presence in the frontier AI market, directly challenging OpenAI and Anthropic. The model's claimed performance improvements and competitive pricing signal an intensified market battle, potentially driving down costs for advanced AI capabilities. Companies relying on AI for complex workflows may find Google's offering a more economical alternative, impacting the pricing strategies of other major cloud providers and AI developers.

The availability of Argon to Pro and Ultra subscribers, with uncertain access for lower tiers, suggests a tiered approach to AI model deployment, aligning with Google's broader strategy for its AI services. This release, close on the heels of Meta's own frontier model announcement, underscores the rapid pace of innovation and competition, indicating that the frontier AI landscape is broader than previously assumed and that the balance of power remains dynamic.