Roblox has launched 4D creation in beta, an AI tool that can generate fully functioning in-game models from natural language prompts, expanding beyond its earlier static object generator.

The company aims to grow its vast developer community and user base—already averaging over 150 million daily active users in Q3—by making creation more accessible across visuals and coding.

Anupam Singh, Roblox’s VP for AI and infrastructure, emphasized the integration of visual tools with programming to spur content creation.

A sample experience, Wish Master, demonstrates the new tools, though early object generation is viewed as more of a novelty than an immediate game-changing feature.

The work remains in a research stage with no published timeline for availability, and reviewers note mixed practicality and skepticism about immediate impact.

Industry observers describe Genie as exciting but not a full game engine; Take-Two cautions that many game-development elements extend beyond world creation.

The broader industry context includes investor reaction to Google’s Genie and skepticism about its capabilities, with Take-Two similarly signaling cautious optimism.

In beta, the tool supports two schemas—drivable four-wheeled cars and single-mesh objects—with examples including cars, a plane, and a gumball gun generated from prompts.

Public demos include a Viking character and a claimed Twin Towers generation, illustrating rapid prototyping but also concerns about output quality and fidelity.

Wish Master serves as a live beta showcase with drivable cars, flyable planes, and animated dragons, gathering user feedback to inform refinements.

Roblox is balancing rapid innovation with safety and regulatory pressures, aiming to expand its creator ecosystem while addressing regional access and safeguards.

Early adopters report strong engagement, with Wish Master users generating over 160,000 objects in six months and average playtime rising about 64%, signaling prototyping potential.