The model: Seedance (ByteDance) — text-to-video that follows numbered, shot-by-shot direction better than flowing prose.
5 starter prompts — NOT TESTED. Every prompt below is written as a numbered multi-shot list, which is how Seedance takes direction: one numbered beat per shot, camera and subject spelled out separately, transitions named instead of implied. But we have not run Seedance ourselves, so unlike our tested Veo set, these pages carry no video, no run log and no receipts — they are starting points to adapt, not results we can vouch for. They were written against Seedance 1.0; newer Seedance versions may name settings differently, so check the suggested settings on each entry against your own panel. When we get run access, this page becomes a tested set with real outputs, generation counts and credits. Until then, if you run one, submit your result: a working run can graduate an entry to Tested, credited to you.


