Use case
Long-form AI video
A single catalog clip is not a five-minute file. Wan 3.0 text-to-video is 2–30 seconds. The studio’s multi-step flow continues from the last frame so a sequence can run to about 5 minutes, which is the limit stated on the homepage.
Open the video studio, generate a clip, then continue from the last frame. Each step spends tokens at that engine’s price.
Questions
Is 5 minutes one generation?
No. It is a flow of clips. One Wan 3.0 text-to-video generation is listed up to 30 seconds.