A Genjutsu prompt only has to say what changes and what must stay. These twelve templates cover the changes people ask for most.
Genjutsu is a reference-driven video model, not a text-to-video model. Your source clip already carries the motion, the camera and the timing, so the prompt never has to describe them. It only has to say which element changes and what that element becomes.
That is why good Genjutsu prompts are short. A long cinematic paragraph leaves the model nowhere to apply your reference images, while a two line instruction about a single element usually lands on the first attempt.
Every prompt on this page is built from the same three parts, in this order:
Higgsfield GenjutsuPut the jacket from image 2 on the fighter. Keep the fight, the camera and the lighting identical.Replace the bottle on the table with the product in image 1. Keep the label readable and keep the hand the same.Change the hoodie to the colour and fabric shown in image 3. Everything else stays the same.Add the sunglasses from image 2 to the driver. Keep his face and the camera move unchanged.Use the woman in image 1 as the lead. Keep every move, the timing and the camera exactly as in the original.Replace the extra on the left with the man in image 2. Keep the crowd around him and the camera path the same.Recast the defender with image 2 and keep the attacker from the original shot. Preserve the choreography.Use the face in image 3 for the fighter. Keep the lighting, the sweat and the dirt from the original.Move the whole fight into the valley in image 1. Keep the choreography and the camera move.Restage the same shot at golden hour. Keep the subject, the wardrobe and the action identical.Add heavy rain to the scene in image 2. Keep the subject dry and the camera unchanged.Make the creature in image 1 three times taller. Keep the street and the crowd the same.Motion Transfer keeps the motion, the camera and the timing, and rebuilds the scene from your references. Prompts for this mode read like casting and staging decisions.
Object Swap changes one selected element and preserves the rest of the original shot. Prompts for this mode are a single instruction: this thing, replaced by that reference.
Attach your references and describe the change in one or two sentences. With no instruction at all the model has no way to know which element you want replaced.
One or two sentences. Name the element, point at the reference image, and say what must stay. Long cinematic paragraphs usually reduce how closely the output follows your references.
The generator takes a single instruction field, so keep the guidance positive. Describe the result you want rather than the artifacts you do not want.
No. The camera belongs to your source clip. Describing a different move in the prompt is one of the most common reasons a recast drifts away from the original shot.
Yes, and it is the fastest way to keep a series consistent. Keep the element wording and the reference images identical, and change only the source clip.
Usually because the prompt asked for too many changes at once, or because one of the named elements had no reference image attached. Split the request and attach an image for every element you mention.
Recast any motion with your characters, locations and products, or swap specific elements and keep the rest of the shot.