Prompt or frame
Begin from a written direction, a first frame or supported references.
AI video production
Turn an idea into motion without losing the project around it.
Start with text, a first frame or visual references. Choose the model that fits the shot, then keep the prompt, source image and finished clip together.
What it gives you
One place to direct, compare and continue.
Soho Frame is built for creators who need more than a single disconnected render. Generate a frame, animate it, review the result and continue the sequence inside the same project.
Curated model choices
Use distinct models for cinematic action, longer scenes, speed or visual continuity.
Project continuity
Inputs, prompts and outputs remain attached instead of disappearing into separate tools.
Compare model guides
Model guides
Generated examples
Real motion made through the studio.
These are generated Soho Frame samples, labeled by the model used. They are not stock promises.
Vintage New York
Wan 3.0
Kingdom Rising
Seedance 2.5
Ink World
Seedance 2.5
A clear workflow
From direction to a usable clip.
- 01
Set the shot
Describe the subject, action, camera and what must stay consistent.
- 02
Choose the route
Pick a model based on duration, references, resolution and audio needs.
- 03
Review in context
The finished clip returns to the project beside its source material.
Questions
AI video generator questions
Can I create from an image?
Yes. Supported models accept a first frame, and some also accept an end frame or multiple reference images.
Which model should I choose?
Use the visible model details and model guides. Duration, resolution, reference limits and audio support are taken from the active Soho catalog.
Where does the result go?
The video is saved to the same project as its prompt and source media, so you can continue instead of rebuilding context.
Direct your first shot.
Open Video, choose a model and keep the result connected to the work that created it.
Start creating