Google opens Genie 4 to developers through the Gemini API

Google made Genie 4 available to developers through the Gemini API and Google AI Studio on November 19, extending access beyond the consumer rollout announced a day earlier.
The official announcement says the model is accessible through the Interactions API. Developers can start with a text prompt, a reference image or both to generate continuous 720p environments running at 24 frames per second.
The generated worlds include synchronized native audio, including ambient sounds, sound effects and character dialogue. Keyboard controls and live text prompts let users navigate and change weather, lighting, objects or events while an environment is running.
Google says Gemini reasoning helps maintain consistent world state. That is the company's description of the system, rather than an independently established reliability guarantee. Outputs include imperceptible SynthID watermarks.
Google AI Studio offers an interactive canvas and starter applications. The developer release matters because it gives builders a route to prototype interactive experiences, spatial designs and environments for robotics or AI-agent training, rather than only exploring worlds inside a consumer product.
Sources and citations
Revision history
Published .