Google Expands Gemini 4 Argon Access, Introductory API Rates Rise After Nov. 30
Google has expanded access to Gemini 4 Argon beyond cybersecurity defenders, making the model available to enterprises and Google AI Ultra and Pro subscribers. The introductory rate starts at $2 per million input tokens and $10 per million output tokens, before increasing after Nov. 30. Google says it has used Gemini 4 for software engineering, cybersecurity and research.
Google has opened Gemini AI model Gemini 4 Argon to enterprise customers and paid subscribers, expanding access beyond the cybersecurity defenders who were initially allowed to use it. The model is available globally through the Gemini app for Google AI Ultra and Pro subscribers, with introductory API pricing due to rise after Nov. 30.
Access starts at $2 per million input tokens and $10 per million output tokens. After Nov. 30, those rates increase to $4 per million input tokens and $20 per million output tokens.
The change gives more businesses and developers access to a model Google presents as useful for demanding technical work. Argon is also available in Android Studio, Google AI Studio and Google Antigravity, the company’s agentic developer platform.
Google cites software and security work
Google says it has used Gemini 4 internally in software engineering, cybersecurity and research. In a company blog post, it said the model helped optimize Python codebases, identify and patch cybersecurity holes, and make its open Gemma AI models three times quicker at tasks.
Google also described using the model in physical AI research, including work on robotics. The company said Gemini 4 helped ground a humanoid device in real-world physics and understand its surroundings.
That is a practical pitch. Developers and research teams could use the expanded access for coding and technical projects, while the price increase gives them a near-term cost change to factor into usage. The source does not specify how much a particular project would cost; charges depend on the number of tokens used.
Google’s focus on technical and scientific applications marks a shift in emphasis from its AI image generator Nano Banana, CNET reported. The company is now highlighting work involving mathematics and science, including robotics and nuclear fusion research.
Access broadens as safety questions persist
The expansion comes as technology companies face criticism over AI, including concerns about data center construction and AI agents making unwanted incursions. CNET described a broader industry effort to stress potential applications such as cancer research as public criticism mounts.
Gemini 4 Argon competes with OpenAI’s GPT-6 Astra and Meta’s Muse Wonder. The outlet also reported that Google had engaged with the White House’s voluntary review process for new models and said Gemini 4 passed inspection on its safety framework. The process is not publicly visible, CNET said.
That assessment does not settle wider concerns about AI safety. CNET noted that worries have grown around “AI doomsday” scenarios over the past year, while the model’s broader availability puts it in front of more paying users and organizations.
For customers deciding when to adopt it, the immediate distinction is price: the introductory rates apply before Nov. 30, after which input pricing doubles to $4 per million tokens and output pricing doubles to $20 per million.
