VIDEO GENERATOR
Three ways into a clip. 23 models to get there.
Start from a sentence, from a still, or from a video you already have. You are not forced to make an image first , that was true once, and it cost people a generation they never needed.
From a prompt
9 models
No source needed. Describe the shot and the motion and you get a clip. Put the movement first , what the subject does, what the camera does , and only then the surroundings.
From a still
12 models
Your image becomes the first frame. Describe what happens next, not what is already in shot , that is on the frame already, and describing it again only adds noise.
From a video
2 models
Take a clip you already have and make a new version of it: a different style, a different look, with the motion left intact. Pick the clip from your library or upload one.
Sound, when the model can do it
Some models generate the audio along with the picture. That setting is the first thing you see, with the price difference next to it, because sound costs more and you cannot add it afterwards. On a model that has no sound it says so , rather than hiding the setting, which just makes the studio look broken.
Your own style, in motion too
A LoRA is not only for stills. Some video models take one, so the character or the look you trained carries into the clip. As on the image side, the studio picks a model that matches the family your LoRA came from.
Long jobs do not block you
Video takes minutes, not seconds. You get a place in the queue straight away and the studio tells you when it is done , you can close the tab. If the provider drops the connection, a second check picks the result up anyway, so a finished clip never sits there looking stuck.
What it costs
Video is priced per second, and the length and the resolution move the amount , which is why it is on the button before you start. Sound is a separate rate on 5 of the models. If something breaks on our side or at the provider you get everything back; if the model refuses after starting, you pay only what the attempt actually cost us.
It connects to the rest
Make the still in the image generator and turn it into a clip in one click, let the assistant work out which model suits the shot, or run the whole series through Flow.