Google has opened Genie 4, its real-time world model, to developers through the Gemini API and Google AI Studio, a day after its consumer launch, Google said on 19 November.
Google’s pricing page puts the effective cost at about $0.10 a second, or ~$6.05 a minute. That is about $363 for an hour of generated world (our arithmetic).
Business Pill · GPUs Explained
An evergreen Business Engineer explainer on GPUs. It teaches the general idea only and does not discuss Genie 4.
The key insight: Genie 4 is now metered for developers at about $0.10 a second of 720p interactive world, roughly $363 an hour on Google’s own effective rate (our arithmetic).
What Developers Get
Starting from a text prompt, a reference image or both, developers can generate “continuous 720p worlds at 24fps with synchronized native audio”, Google says.
Real-time keyboard controls and live text prompts let users “change weather, lighting, objects, or events on the fly”. Every output carries SynthID watermarking, the post says.
Access runs through the Interactions API on the paid tier only. The pricing page says Genie 4 does not support the Free tier, Batch API, Flex inference or Priority inference.

How the Meter Works
Input costs $1.50 per million tokens for text or a reference image. Video output, with audio included, costs $17.50 per million tokens.
Billing counts 240 tokens per frame, or 5,760 tokens a second at 24fps, the page says. Google itself converts that to “approximately $0.10 per second (~$6.05 per minute)”.
Sessions bill until they end: the page tells developers to “Always call complete() to end an active session and stop billing”.
The Numbers in Proportion
At ~$6.05 a minute, an hour of continuous Genie 4 output costs about $363, and a 3-minute run about $18 (our arithmetic).
Consumers reach Genie 4 through Project Genie for Google AI Ultra, which Google’s plans page lists from $99.99 a month. That monthly fee buys about 16.5 minutes of API output at the effective rate (our arithmetic; the consumer plan’s usage limits differ).
The Structural Read
As we read it, per-second billing makes world models a variable cost like video generation, not a flat subscription feature.
As we read the session rule, cost control shifts to developers: an open session keeps billing until complete() is called.
Google, Gemini API pricing page
“approximately $0.10 per second (~$6.05 per minute)”
Three Implications
FOR DEVELOPERS Genie 4 is paid tier only, with no Free, Batch or Flex option, per Google.
FOR ROBOTICS AND AGENTS Google pitches it for prototyping and for training robotics and AI agents.
FOR CONSUMERS Project Genie stays the consumer route, through the Ultra plan.
What Is Not Established
Google has not said how much Genie 4 usage an Ultra plan includes, or whether API rates will change after this launch. Our hourly figure assumes uninterrupted output at Google’s quoted effective rate.
The Bottom Line
Genie 4 is now a metered developer product: about $0.10 a second of interactive 720p world, roughly $363 an hour on Google’s own effective rate (our arithmetic).
97,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. Availability, features and watermarking are from Google’s 19 November blog post. Token prices, tokens per frame and the ~$6.05-a-minute effective rate are from Google’s Gemini API pricing page, read on 19 November. The Ultra plan price is from Google’s plans page as quoted in our 18 November piece. Hourly, 3-minute and minutes-per-plan figures are our arithmetic.
Sources: Google, ‘Genie 4 is now available on the Gemini API.’, 19 Nov 2026 · Google, Gemini API pricing page, Genie 4 section, read 19 Nov 2026 · FourWeekMBA, ‘Google Opens Genie 4 to Ultra Users 22 Days After Its Debut’, 18 Nov 2026









