Clip Length Arbitrage

Clip Length Arbitrage is the practice of pricing text-to-video work by the usable second rather than by the render or the project. Most engines in this category output short bursts: MiniMax H3 produces 4 to 15 second clips with native audio and lip-sync, and MiniMax H3 Max works in a 5 to 15 second window with camera-motion and character-consistency controls. That ceiling is not a limitation to apologize for, it is the unit of economics. A 30 second social cutdown is three to six renders plus selection and stitching, so the agency's real cost sits in the selection pass, not the generation. Agencies that quote a flat creative fee absorb that variance; agencies that quote per finished second, with a defined revision allowance, protect margin while clients get predictable pricing. AIProductAds shows the same math at 2K with synced audio, where a single 15 second ad variant is the deliverable unit.