VidModelHub

Free tool · No signup · 5 plans/day

Shot Planner — from idea to shot list

Every AI video tool gives you a blank box. Directors don't start with a blank box — they start with a shot list. Describe what you're making; get back numbered shots with camera moves, ready-to-generate prompts, and the real cost of the whole piece — before spending a credit.

Total length

Style (optional)

Free · 5 plans per day · powered by an LLM tuned on the official Seedance prompt formula

Why plan shots before generating?

Because wasted generations are where AI video budgets die. A 30-second piece generated as one prompt gives the model too many jobs at once; generated as six planned shots, each prompt does one job well, failures are cheap to re-roll individually (draft at 480p for 2 credits), and the cut has actual rhythm. This is how production works — we just compressed the pre-production into ten seconds.

Frequently asked questions

What does the shot planner actually do?
It turns a one-line brief into a director's shot list: 3-8 shots with durations, camera moves, and a complete generation prompt per shot following ByteDance's official three-block prompt structure (subject & action → environment & style → camera & audio). It also totals the exact credit cost of generating the whole plan before you spend anything.
How does it keep characters consistent across shots?
The planner locks one precise descriptor phrase per recurring subject and repeats it word-for-word in every prompt — the most effective consistency technique available without reference images. When generation opens, you can add reference images on top for even tighter consistency.
Is it really free? What's the catch?
5 plans per day, free, no signup. The 'catch' is honest: we hope you'll generate the shots here when Seedance 2.5 opens (first clip free, then one-time credit packs), and the optional email-me-this-plan button subscribes you to our launch alerts.
Which models do the prompts work with?
They're written for the Seedance family's prompt conventions, which transfer well to Kling and Veo with minor tweaks. Multi-model prompt conversion is on our roadmap.