How to Make a Chuttamalle AI Video from Two Photos

Prepare two photos before you begin

Choose one clear photo of each adult you want in the clip. You need permission to use the photographs and permission from both people to create the proposed scene. Avoid a photo with several faces, a face hidden behind an object, or a heavily filtered selfie. A visible face and upper body help establish identity and clothing without forcing the model to invent everything from a tiny headshot.

Keep the references reasonably similar in lighting and clarity. A sharp daylight portrait paired with a blurry night photo gives the model uneven information. If possible, use uncomplicated backgrounds and leave a little space around the person. You do not need an original dance video, and uploading one is not part of this workflow.

1. Open the creator and sign in

Open Chuttamalle AI and sign in before submitting a generation. Check that the service is available and that your balance covers the displayed credit cost. The current two-photo generation requires 20 credits and produces an eight-second clip. If the provider is not configured, the creator cannot produce a video; repeated submissions will not resolve that availability issue.

2. Upload one reference for each person

Place the first person’s photo in the first upload slot and the second person’s photo in the second. Check the previews before continuing. If you accidentally choose the same image twice, replace the incorrect one. Follow the upload control’s file type and size instructions; a renamed file extension does not change the actual image format.

Choose 9:16 for a vertical composition or 16:9 for a horizontal one. Consider where the finished clip will appear. Both people need room in the frame, so a vertical composition may place them closer together. Changing the crop after generation can remove important movement near the edges.

3. Generate and wait for the result

Submit once and wait while the task processes. The model creates an original dance scene using the two appearance references. It is not an exact recreation of a filmed performance, and it does not include the original song. Avoid starting several attempts just because the first one is still processing; each accepted generation can consume credits.

When the result is ready, play the entire clip before downloading. Look for changes in faces, awkward fingers, clothing inconsistencies, and whether both subjects remain visible. Keep a copy of the result you choose to use. If you plan to add music, follow the separate soundtrack workflow.

If something goes wrong

• Upload fails: check the allowed format and size, then export a fresh image from your photo editor. Confirm your internet connection before trying again.
• Not enough credits: review your balance and the creator’s current cost. Do not assume a failed submission purchased or added credits.
• Generation fails or stays pending: keep the task information and check the status before submitting another attempt. Ask support to investigate if the status remains unclear.
• The video looks wrong: improve the references, reduce ambiguity in the scene, and evaluate whether another paid attempt is worthwhile. A new generation is not guaranteed to correct every defect.

For unresolved issues, sign in and open a support ticket with the task reference and a concise description. Do not include passwords or payment card details. Read what Chuttamalle AI videos are for the differences between a generated clip, an existing performance, and a song edit.