Inputs
One-frame and start-end workflows
Standard image-to-video starts from one reference frame. When the selected workflow exposes a second frame input, the model can instead generate a transition between a chosen start and end state; use only the input slots and controls shown for that workflow.
Examples
When this article is useful
Animate a permitted reference image with NSFW-friendly AI video workflows. Review input requirements and credit cost before generating private clips. 18+.
A focused first test
Standard image-to-video starts from one reference frame. When the selected workflow exposes a second frame input, the model can instead generate a transition between a chosen start and end state; use only the input slots and controls shown for that workflow.
Failure modes and fixes
Limitations and decision check
Acceptance test: use this reviewed requirement as a pass/fail criterion on one representative result; compare the result with the requirement and stop, revise, or choose another route if it is not met: Do I need more than one photo? Standard image-to-video uses one reference image. Some workflows expose a second start-end frame input; if that option is available, Studio shows both required slots before generation.
Motion quality is decided across the whole clip, not by the first frame. Inspect face and limb continuity, camera direction, background drift, and the final frame before accepting a render or extending it.
