Google Opens Genie 4 World Generation to Gemini API Developers

Unite.AI
Read full story· 4 min
Google launched Genie 4 on November 19, 2026, via the Gemini API and Google AI Studio, enabling developers to create interactive 720p video worlds controlled by keyboard and text prompts with synchronized audio.

What happened

On November 19, 2026, Google released Genie 4 to developers through the Gemini API and Google AI Studio. This launch allows developers to create interactive 720p video worlds that respond to keyboard controls and text prompts, with synchronized audio. The API supports seeding worlds with text or images and enables real-time user interaction. The release follows a prior update that gave early access to certain Google AI subscribers.

Key details

  • Genie 4 outputs 720p video at 24 frames per second with stereo audio.
  • Sessions are capped at 120 seconds, with the most stable visuals in the first 45 to 90 seconds.
  • Only paid-tier Gemini API users can access Genie 4; it is not available on the free tier.
  • Session limits vary by usage tier, with Tier 1 allowing 2 concurrent sessions and 50 requests per day.

Why it matters

Developers using the Gemini API can now build interactive video experiences that respond in real time to user input, but must manage costs and session limits. The requirement for a paid API key and the absence of a persistent world mean applications are limited to short, non-continuous simulations. This affects companies and developers seeking to prototype interactive environments, as they must account for pricing, session duration, and technical constraints when integrating Genie 4 into their products or research.


More in Video Generation