GenSpicy
AI template

AI Couple Video from Two Photos

Add one photo of each person, pick a moment such as a reunion hug or a gentle kiss, and get a short, clothed romantic video.

Person 1 photo

One adult, face clearly visible, ideally waist up. Use your own photo or one of someone who agreed.

Person 2 photo

One adult, face clearly visible, ideally waist up. Use your own photo or one of someone who agreed.

Use one clear, front-facing photo per person, waist up, with nothing covering the face.

Write “Person 1” and “Person 2” to say who does what.

0/240
0/200
0/160
0/300
Length
Format
Model
Add the required inputs to see the price.

Cover illustration

Add one photo for each person: Person 1 and Person 2 are sent to the model as separate references, in that order. Pick a moment, a setting and a camera style, or describe your own. Faces and outfits follow your photos as a guide, so they can still come out a little different.

How it works

  1. 1Add a photo of Person 1 and a photo of Person 2.
  2. 2Pick a moment, setting and camera, or write your own details.
  3. 3Choose a model, check the live price and generate.

Add the two photos separately

Put one adult in Person 1 and the other in Person 2. You don’t need to combine the photos first. MiniMax H3 and Wan 3.0 both use the two uploads as separate video references, in that order.

Choose clear, front-facing photos taken from around the waist up. Similar lighting and camera distances help the pair fit into the same scene. Avoid sunglasses, deep facial shadows and crowded backgrounds. Use photos of yourself or adults who agreed to the video, with both people fully clothed.

Pick a moment and a place

Moment, Setting and Camera each have their own field. They start empty: leave one blank and a default is used, or pick a chip or type your own. Choose a reunion hug, holding hands, a forehead kiss or a slow dance, then add a place. An airport arrival works for a reunion; a café or beach can suit a quieter moment. Keep the scene in one location so the clip has time for the interaction.

Use Person 1 and Person 2 in your details when the roles matter. For a reunion, say who approaches and who waits. For a walk, say whether they look ahead or turn toward each other. A clear action tells the model more than a string of words such as romantic, beautiful and cinematic.

Choose the length and framing

Both models offer five or ten seconds, with portrait 9:16, landscape 16:9 and square 1:1 formats. Portrait gives you a vertical clip. Landscape leaves more room beside the pair if the surroundings matter. Ten seconds gives the scene more time, though it won’t guarantee better faces or more accurate movement.

Match the camera to the action. A close view suits a forehead kiss; a wider shot lets you see a walk or dance. If you ask for a quiet embrace, keep camera movement gentle too. Several turns, a sweeping orbit and a close-up can be too much to fit into one short video.

Check the whole clip before sharing

Watch both faces as they move, then check hands, overlapping arms and the point where the people touch. Faces, outfits and relative height can drift, even when the beginning looks convincing. Exact choreography and physical contact aren’t guaranteed.

Your completed video is saved in your private library, where you can review and download it. If a hug looks awkward, try a simpler interaction on the next run. Change one choice at a time so you can see what helped.

Tips for the best result

  • Similar lighting in both photos helps the two people look like one scene.
  • Use “Person 1” and “Person 2” in your details to say who does what.
  • Faces and clothing can come out slightly different from your photos, so check both people before you save or share a version.

Questions

More templates