Raindance Maker

How to Make a Raindance AI Video from Two Photos

Upload two photos with one person each. Photo 1 replaces the person seated on the pier at the start; Photo 2 replaces the person who sings most of the close-ups. Pick a Seedance model, a length from 5 to 30 seconds and a resolution, check the credit total, then download a 4:3 MP4 with the reference soundtrack already added. No prompt, filming or editing app is needed.

Facts last checked: 2026-10-04

What you need

  • Two photos, one person in each. They can be taken at different times and places.
  • An account to generate and keep your videos.
  • Credits for the length, model and resolution you choose. The generator shows the exact total before you start.

You do not need a prompt, a green screen, video editing software or a copy of the song.

Step 1: Choose a photo for each role

The template has two fixed roles. The role follows the performer through every cut, not a side of the screen:

Person 1 · starts seated on the pier
Person 1 · starts seated on the pier
Person 2 · walks in and takes the close-ups
Person 2 · walks in and takes the close-ups
  • Person 1 starts seated at the end of the pier and appears mostly in the wide and shared shots.
  • Person 2 sings to the camera and carries nearly all of the close-ups. Give this role your clearest, sharpest face photo.

If you upload them the wrong way round, press Swap photos before generating. The shot-by-shot breakdown shows how much screen time each role gets.

Photos that work best

  • One person per image, face visible, front-facing or a slight three-quarter angle.
  • Waist-up or full-body shots give the model more information about clothing and build.
  • Even light, no sunglasses, masks, hands over the face or heavy beauty filters.
  • JPG, PNG or WebP, up to 20 MB each, shorter edge at least 300 px.

Only use photos of yourself or of people who have agreed to take part.

Step 2: Pick the length, model and resolution

SettingOptionsWhat it changes
Length5, 10, 15 or 30 secondsHow much of the performance you get. 30 seconds covers the whole remake.
ModelSeedance 2.0 Mini, 2.0 Fast, 2.0, 2.5Cost and rendering approach. Mini, Fast and 2.0 join two 15-second parts for 30 seconds; 2.5 makes one continuous clip.
Resolution480p, 720p; 1080p on 2.0 and 2.5640 × 480, 960 × 720 or 1440 × 1080, always 4:3

The generator defaults to Seedance 2.5, 10 seconds at 720p. Every combination has a fixed credit price, shown in the model menu and below:

Video length
Raindance AI Video Pricing
Model10 seconds · 480p10 seconds · 720p10 seconds · 1080p
Seedance 2.0 MiniCosts 11 creditsCosts 24 credits—
Seedance 2.0 FastCosts 40 creditsCosts 87 credits—
Seedance 2.0Costs 68 creditsCosts 145 creditsCosts 364 credits
Seedance 2.5Costs 97 creditsCosts 219 creditsCosts 388 credits

We do not rank the models for likeness: no model is guaranteed to keep both faces perfect across every cut, so review the result whichever you choose.

Step 3: Generate, review and download

  1. Press Generate Raindance video. Your photos upload for this video only, and the credits are charged once.
  2. Wait for the progress bar. The model renders a silent video first; a finishing step adds the soundtrack from the same reference clip.
  3. Play the result with sound. Check both faces in the close-ups, the shared shots on the sand and the wide pier shot.
  4. Download the MP4. Your recent Raindance videos stay in the history under the generator.

Ready to try it? Upload one photo for each role and check the credit cost before you generate.

Make my Raindance AI video

Making it on your phone

The whole flow runs in a mobile browser: pick the photos from your camera roll, generate and download the MP4 to your phone. No separate app is needed.

Posting to TikTok, Reels or Shorts

  • The video is 4:3, so vertical apps show it with bars above and below. Don’t crop it to 9:16 or you will cut off a performer.
  • The download already contains the soundtrack. Platforms may mute or claim music in some regions; their rules apply.
  • Label the post as AI-generated. TikTok explains its AI-generated content labels.

If something goes wrong

ProblemWhat to do
The roles are the wrong way roundGenerate again with the photos swapped; a new generation is charged again.
A face drifts back to the original performerUse a sharper, front-facing photo for that role, or try another model.
The photos were rejectedUse photos without nudity, minors or violent content, one adult per photo.
The generation failedIts credits are returned automatically. Try again or contact support.

Ready to try it? Upload one photo for each role and check the credit cost before you generate.

Make my Raindance AI video

FAQ

Can I make a Raindance AI video with only one photo?

Not on Raindance Maker. The template has two roles, so it needs one photo for each person. There is no solo mode.

Can I get a 9:16 vertical video?

No. The output is 4:3 so both performers stay in frame. You can post it to TikTok or Reels as is; the app shows bars above and below.

Do I need to write a prompt?

No. The role instructions and the reference performance are built in. You only choose the photos and settings.

How long does it take?

Most videos finish in a few minutes; longer clips and higher resolutions take longer. You can leave the page and find the video later under My videos.

What if a face does not look like the photo?

Try a clearer, front-facing photo for that role, especially for Person 2, who has the close-ups. Likeness can drift across cuts with any model, so review the whole clip before sharing. Failed generations return their credits; a finished video with imperfections is not refunded automatically.