CVCOVANewsComputing, Opportunities, Value & Analysis
Artificial Intelligence3 min read

DeepMind shows Gemini 4 Argon building a 3D Gothic cathedral with Blender MCP in Antigravity

Google DeepMind posted a video on X in which "Gemini 4 Argon" builds a 3D Gothic cathedral from a single Nano Banana image, using Blender MCP inside Antigravity. It is a demo: the material does not mention the model's availability, pricing or specifications.

By COVA News newsroom
Temple and Convent of Santo Domingo de Guzmán (Centro), Puebla de los Ángeles, State of Puebla, Mexico

Photo: Catedrales e Iglesias · CC BY 2.0 · via Flickr

Video · YouTube

On October 23, 2026, Google DeepMind's official X account posted a short video showing a model called Gemini 4 Argon generating a 3D scene of a Gothic cathedral.

What DeepMind showed

According to the post, the workflow starts from a single image created with Nano Banana. From there, using Blender MCP, Gemini 4 Argon built the cathedral inside @antigravity. DeepMind says the model "iterates on geometry, textures and lighting" until it reaches the finished scene.

Summary of the elements mentioned:

  • Starting point: a Nano Banana image.
  • Connection to the 3D tool: Blender MCP.
  • Environment where it runs: Antigravity.
  • Result shown: a 3D Gothic cathedral, with geometry, textures and lighting refined through iteration.

Why it matters

The demo points to AI being used as an agent that operates 3D modeling software and refines its own output step by step, rather than delivering an image or a file in a single attempt. If this works outside the video, it could shorten visual prototyping tasks for people working in 3D, games or visualization.

For now, this is a reading of what the clip shows, not a verified conclusion.

What we don't know yet

  • whether Gemini 4 Argon is available, and to whom;
  • release date, country or pricing;
  • license or access restrictions;
  • how long the scene took or how many attempts it required, or whether the video is edited;
  • results from independent testing or comparisons with other models.

Until DeepMind publishes documentation or a product page, the scene is best treated as a company demo rather than a confirmed capability available to users.

Source: Google DeepMind on X

Related news

Artificial Intelligence

Mathematician says he has proved the Riemann hypothesis with help from Claude; it still needs to be verified

Levent Alpöge has published a purported proof of one of the most famous problems in mathematics, produced together with Anthropic's AI. Early reviews are encouraging, but the community has not yet validated it.

Artificial Intelligence

OpenAI cuts GPT-6.2 Astra API pricing by 20%: what changes for those paying in dollars

OpenAI launched GPT-6.2 Astra and GPT-6.2 Sol. The flagship model's usage-based pricing drops, but the cheaper model's stays the same, and the fine print will determine how much users in Latin America actually save.

Artificial Intelligence

Nvidia reports earnings without Jensen Huang: what's at stake for AI chips

Five days after its founder's resignation, the company that dominates artificial intelligence chips faces Wall Street. For Latin America, open questions outweigh measurable effects for now.