Choose a better reference photo, understand cartoon limitations, and troubleshoot face drift or unnatural motion.
One recognizable subject per photo gives the generator a clearer reference. Check the input before starting a render.
Review the preview before buying. A separate paid render can differ, and a better photo cannot guarantee a perfect face or motion. Read how photos are processed.
Straight answers about this website. Browse all questions.
The current modes are designed for photographs of adults or pets and aim for a photoreal cafe scene. There is no dedicated cartoon or illustration mode, and we cannot promise that an illustrated character or its art style will be preserved. For the most predictable input, use a supported real photo you have permission to use. Do not buy a pack expecting a guaranteed cartoon conversion.
A small, blurry, filtered, shadowed, or side-facing reference can make identity harder to preserve. Multiple faces or the wrong Duo upload order can also confuse the result. Try a sharper, front-facing photo with even light and one subject per file, check the roles, and use a free preview when available. AI may still change facial details. A completed video with a likeness concern is not automatically a failed render; contact support with your order reference if needed.
Start with an unobstructed face, natural proportions, even light, and a simple background. Use a separate photo for each Duo role and avoid extreme close-ups, heavy beauty filters, motion blur, or cropped facial features. Pet photos should show a single animal clearly. Review the preview for face drift, hands or paws, cups, and the table transition. These steps can help, but no input can guarantee artifact-free AI motion.
One photo. Free 5s preview. Full 15 seconds from $9.90.
Make my fancy cafe video