All guides

Getting started

How to make a short AI rap video from your photos

Choose a photo mode, write a short verse, preview the layout, and follow your selected render through to an MP4 download.

Start with photos you have permission to use and a short original verse. Choose two separate adult portraits, one photo of two adults, a solo adult portrait or a pet photo. The studio offers 5-, 10- or 15-second requests in Standard 480p or HD 720p and shows the quote and availability before you confirm. These are generation controls; identities, pronunciation and movements still need review in the finished result.

1. Choose one idea for the performance

Keep the topic small enough for one short exchange: a birthday toast, a shared late-night story, or a friendly challenge. Decide what the two people are celebrating or responding to before choosing rhymes. You can explore birthday and friend ideas or use the lyrics tool to prepare an editable draft.

Start with a few concise lines rather than a full song. The generator is asked to perform suitable lines from your text; it cannot promise to include every word or assign each line to a particular person. Read the draft aloud and shorten anything that feels rushed.

2. Choose a photo mode

Open the video studio and choose the mode that matches your files: two separate adult portraits, one shared photo of two adults, one solo adult portrait or one pet photo. Choose clear, well-lit references without heavy filters or objects covering important features. In the two-portrait mode, matching backgrounds are unnecessary; the selected stage supplies the performance setting.

Use JPG, PNG, or WebP files, each no larger than 10 MB. Each image must be at least 301 pixels and no more than 5,999 pixels on either side, with a width-to-height ratio greater than 0.4 and less than 2.5. Resize a very tall screenshot or panoramic crop before trying it. See the photo checklist for practical selection advice.

The selected mode determines whether you need one or two photos. Do not include minors in any mode. Obtain consent from the adults shown, and use a pet photo you have permission to use. The model may interpret a shared photo or pet reference differently from your intention; these modes do not guarantee exact identity or behavior.

3. Set the stage, style, and frame

Choose Orange Booth, Street Cypher, or Studio Session. Then select Freestyle, Trap, Storytelling, Diss, or Melodic as the performance direction. These choices guide the output; they do not select a famous artist's voice or guarantee a particular delivery.

Choose a vertical 9:16 or landscape 16:9 frame to suit where you intend to use the video. Check the free layout preview before paying. It helps you prepare the composition and lyrics, but does not animate the portraits or preview the final lip sync. Generated vocals are requested in English. Alternatively, the own-audio option accepts a permitted MP3 or WAV reference, 2–15 seconds and up to 15 MB, long enough to cover the selected video duration. A file used only for local listening is not uploaded. An uploaded audio reference guides generation; it does not guarantee unchanged original sound, exact lip sync or voice cloning.

4. Review the cost before submitting

Select 5, 10 or 15 seconds and 480p or 720p, then check the displayed credit requirement. At 480p those lengths use 4, 7 or 10 credits; at 720p they use 8, 15 or 22. Sign in when requested, review the confirmation, and submit once when rendering is available. Previewing the layout does not spend video credits. A new completed version is a separate render request, so changing your mind and rendering again uses the displayed credits again.

Read credits and failure handling before purchasing if you are deciding between a one-time pack and a monthly plan.

5. Retrieve and check the finished MP4

Open My videos with the same account to follow the task. When it succeeds, play the result and use Download MP4 to keep a local copy. Check the faces, pronunciation, motion, and first and last frames before sharing. AI generation can vary; the layout preview does not guarantee the final performance.

If a task is still waiting, avoid submitting it again just to see whether it works. Use the result and download checklist to distinguish a pending task, a confirmed failure, and a playback problem.

Prepare your photos in the studio →