Try it

  1. Define each role, such as images for the subject and texture, video for camera work and movement, and audio for rhythm.
  2. State each asset's role in the prompt too.
  3. Check that the input mode supports the combination before generating.
Example prompt
Use Image 1 for the mug design and lighting. Use Video 1 for the slow camera movement.

Requirements and considerations

For example, Runway's MiniMax H3 lists support for multiple reference inputs. However, its keyframe and reference inputs cannot be used together. Check each model's limits and supported combinations.

Official guides and references