Start with clear, authorized photos of both people. If your chosen Media.io model supports only one input image, combine them into a single image first. Describe the cliff setting, each character, camera movement, and sequence of actions, then generate and review the result.

Try: “Animate the two people in the reference image on a windy coastal cliff at dusk. They exchange a tense look as the camera slowly circles. One steps back dramatically; the scene ends with a wide cinematic shot and a non-graphic slow-motion fall. Preserve both faces and outfits, realistic movement, stormy lighting, no gore, no extra people.” For complex plots, split the story into separate shots.

You can make an original fictional variation in which the apparent power dynamic reverses at the cliff. Use distinct character descriptions and a controlled sequence of shots. Exact replication of another creator’s template, footage, or soundtrack may require permission, and results will vary by model.

Media.io model availability, credits, output length, and pricing depend on the selected workflow. An 18-second example from another provider does not guarantee the same duration on Media.io. If the model supports shorter clips, generate multiple segments and combine them in an editor.