Editing Openai/69557275-0da8-8004-a03a-8eb02e10a61f (section)

===== Your kid asks for a 10-second video: “Make our dog a knight battling a dragon.” =====

Your AI replies like a utility company:

“Okay. Standard quality: ~$1. Pro quality: ~$3–$5.”
(Those are literal per-second prices in today’s video APIs: ~$0.10/sec for baseline, up to ~$0.30–$0.50/sec for higher tiers.) OpenAI<ref>{{cite web|title=OpenAI|url=https://openai.com/api/pricing/|publisher=openai.com|access-date=2026-01-02}}</ref>

You pause. Not because you can’t afford it—because it changes the vibe of imagination.

Text will be cheaper, but it’ll still have a meter.

A normal “good assistant” chat might be a couple thousand tokens in and out. On current flagship pricing, that’s pennies; on cheaper models, fractions of a cent. OpenAI<ref>{{cite web|title=OpenAI|url=https://openai.com/api/pricing/|publisher=openai.com|access-date=2026-01-02}}</ref>
But scale it up—long drafts, huge documents, hours of voice, lots of images/video—and you’ll suddenly see why families end up saying things like:
* “Do a cheap pass first.”
* “Run it mini unless it’s important.”
* “Save the big model for the lawyer email.”
* “We’re not burning video tokens on a joke.”

Not dystopian. Just… new.

Intelligence becomes a utility: abundant, magical, and not quite free. OpenAI<ref>{{cite web|title=OpenAI|url=https://openai.com/api/pricing/|publisher=openai.com|access-date=2026-01-02}}</ref>