Google is now rolling out support for Gemini 4 Argon to Antigravity, its agentic AI coding tool, for AI Pro and AI Ultra subscribers. Gemini 4 Argon is Google's latest flagship AI model, announced at the end of September and widely rolled out to users earlier this week. The integration into Antigravity was announced earlier this week and is now live, as confirmed by 9to5Google.
Antigravity is an agentic development platform that uses AI and agents to assist with coding and building apps. With Gemini 4 Argon, Google says the model is 'designed to tackle your hardest problems.' Subscribers get a separate weekly-only Argon quota on top of the existing quota for Gemini 3.8 Flash, allowing them to choose the model that best fits their needs.
Gemini 4 Argon Capabilities and Robotics Applications
Beyond coding, Google is positioning Gemini 4 Argon as a versatile model for robotics tasks. According to The Decoder, Argon controls robots, generates simulations from photos or sketches, and evaluates training videos. On Google's own ERQA benchmark, it scores 87 percent, ahead of other frontier models. However, published example prompts show the demos still require significant human guidance.
In one demo, a Boston Dynamics Spot robot practices tossing soda cans into a bin by first building a virtual simulation. Argon uses simulations to prepare tasks, verify them, or generate training data. From a single photo, it built a virtual replica and generated over 4,500 training episodes, improving real-world success rates from 56 to 75 percent and reducing execution time from 35 to 24 seconds.
- Controls robots and generates simulations from photos or sketches
- Scores 87% on Google's ERQA benchmark
