Available now

Vidu Reference to Image: image from references

By ShengShu

Vidu Reference to Image makes a new still image from your prompt, using reference images to keep a subject's look the same.

Model ID
Maker
ShengShu

Input

Fields marked * are required. Review the price before generating.

Describe the image to create, up to 1,500 characters.

Example prompt. Edit it or write your own.

The reference images to use for consistent subject appearance

Uploading a file needs an account. Create account

JPEG, PNG, WebP · up to 25 MB

Additional settings

Random seed for generation

Estimated total$0.08

You sign up first. Your settings are kept.

About this model

Overview

Upload one or more reference images of a person, character, product or place and describe the new picture. The subject keeps its appearance from the references while the scene, pose and action follow your prompt. The output is 16:9, 9:16 or 1:1.

What you can build

  • One character shown in many different scenes
  • Product shots in settings you describe
  • Consistent key art for a series or campaign
  • Start frames for a video that keep the subject consistent

Prompting tips

  • Use clear references where the subject is fully visible and not covered by other objects.
  • Say in the prompt who or what from the references should appear and what they are doing.
  • Keep the seed to repeat a result in another aspect ratio.

What to check

Compare identifying features, clothing and proportions against the references; inspect overlaps between subjects.

Inputs

  • Prompt (required): Describe the image to create, up to 1,500 characters.
  • Reference Images (required): URLs of the reference images to use for consistent subject appearance.
  • Aspect Ratio (optional): The aspect ratio of the output image. Options: 16:9, 9:16, 1:1.
  • Seed (optional): Random seed for generation.

Output and limits

  • Aspect ratios: 16:9, 9:16, 1:1.
  • Small or dense text in the image can come out misspelled.
  • Hands, crowds and complex poses can show artifacts; generate a few variations and pick the best.
  • Some prompts or outputs can be refused by safety checks.

Price guide

  • 1 image: $0.08 / image