Gemini 4 Argon launches: Google demonstrates 3D scene creation with Blender

Google DeepMind made Gemini 4 Argon available globally on 13 October 2026. Developers and businesses can access the model through the API on AI Studio and the Gemini platform. Google AI Ultra subscribers can use it in Antigravity, while Google AI Ultra and Pro subscribers can use it in the Gemini app and AI Mode in Search. The announcement also included examples of the model being used to generate 3D scenes with Blender.

Google’s 3D creation examples

Google says Argon created a detailed temple scene featuring multiple structures in the Antigravity environment. According to the company, Blender was used in the production process; the resulting scene includes vegetation and lighting as well as architectural elements.

In another demonstration, the starting point is a single Nano Banana image. Google says Argon processed the image through Blender MCP and transformed it into a Gothic-style scene. The process also involved scene decisions such as camera placement and lighting, according to the company’s description.

The company says Antigravity was also used to create a game involving steps such as modeling, rigging, and animation with Python scripts. In this example, the work is brought to the web using Three.js and GLSL shaders. Google also says it used Argon for complex 3D visualizations and interactive animations that can run in a browser.

These are application examples presented by Google as part of its launch promotion. The announcement does not include a detailed technical breakdown of how these workflows could be used in architectural projects, the accuracy of the generated geometry, or Blender version compatibility.

Access requirements and introductory pricing

The API is available to developers and businesses through AI Studio and the Gemini platform. A Google AI Ultra subscription is required to use Antigravity. For the Gemini app and AI Mode in Search, Google AI Ultra or Pro is specified. The announcement does not state any access conditions for free users.

The API prices announced by Google apply during the introductory period: $2 per million input tokens and $10 per million output tokens. The company also says cached input tokens will receive a 95% discount. These figures were announced as introductory prices, not as permanent rates. Pricing details for app and Antigravity access are subject to separate subscription terms from the token rates listed here.

Impact on architectural and visualization workflows

For architects, interior designers, and visualization teams, the main takeaway is the opportunity to try scene generation from text or image inputs, alongside script-based tasks, within the same workflow. Google’s examples point to possibilities such as developing concept scenes, creating 3D content, and automating some production steps. However, the demos do not prove that every project will yield accurate dimensions, clean geometry, or production-ready files.

Before adopting a workflow, teams should evaluate the type of access, usage costs, and compatibility with existing tools. API usage charges input and output tokens separately, while Antigravity and the Gemini app are subject to subscription terms. Google did not specify local hardware requirements or Blender version support in the announcement. Teams should therefore start with a small test to assess scene quality and suitability for their workflow.