Understand video models, durations and frames
Read model-specific requirements without assuming all video engines work the same way.
Select the model before writing the final brief
The chosen model determines clip duration, resolution, audio support and reference requirements. A capability listed for another model does not carry over. The model selector and settings panel are the current in-app reference.
Start frames, end frames and references
A start frame anchors the opening image. An end frame specifies the target final image when supported. A reference guides appearance or motion according to the selected mode. Upload every field marked required; do not substitute a text description for a required file.
Duration is a constraint
Pick from the durations actually offered for that model. Do not write a 30-second sequence into a 5-second clip. For longer storytelling, plan separate shots and review each one. Available resolution, sound and multi-shot options vary by model and can change the credit quote.
Audio and multi-shots
Only request generated sound through models and settings that support it. Multi-shot settings are not present on every model; when offered, keep per-shot prompts and durations coherent. A model’s name alone is not a promise of audio, lip sync or arbitrary editing.
Unavailable does not mean your account is broken
Models can be unavailable due to provider conditions or pending support. Select another active option if the tool reports it cannot run the request. If an accepted job fails, inspect its recorded state before resubmitting.