Wan 2.7 Pro image-to-image edits or composites one to nine reference images under a single instruction, at a 2k canvas of about 4 megapixels. It is built to hold compound, multi-part edit prompts, and this endpoint is uncensored.
Filtering differs by endpoint: Image editing — Uncensored, Text to image — Softened.
Workbench
These examples were written for this endpoint. Replace the prompt and sample media URLs with your own accessible inputs, then check the current quote and model availability before submitting.
Outputemerald-gown-balcony Three references in one edit: a face from image 1, an emerald satin gown from image 2, an art deco balcony from image 3. The instruction gives each image its job in plain words, then adds the pose and the light: a look back over the shoulder, sconce light on the satin.
Outputglass-onto-marble The `prompt` gives one job per image: lift the glass from image 1, place it on the marble from image 2, matching its light direction and colour temperature, glass itself untouched. The result keeps the exact condensation and ice cube from the reference glass, set into warm counter-top sunlight with no stray shadow. A fixed `seed` makes it easy to swap a new object into image 1.
- 118834
Outputoutfit-and-sign Two unrelated edits in one pass: a new outfit on the woman, and new words on the neon sign behind her. The lettering has to come back legible in the same pink neon, at the same size, while her face, her pose and the room stay exactly as they were.
This example starts at $0.075 per image. The Playground updates the estimate when you change the settings.