How to Make the Raindance AI Trend with Your Own Photos (2026)

Understand the scene first

The Raindance AI two-person idea refers to a person waiting on a wooden dock while another sings, followed by the pair together at the beach at sunset. The name points to the Dave and Tems song, but the scene requested by this site has no actual rain. This generator asks for a new short video from your photos. It does not insert shots from the original music video or supply the original song.

Decide who should wait and who should sing before you choose files. The main generator is a two-person sequence, not a single-portrait singing filter. Its requested shots include the dock, a singing close-up and a sunset beach moment. These are directions to the model; the actual sequence, faces and cuts may vary. There is no reference clip to upload and no scene prompt to write for this preset.

Make it in your browser

1. Prepare two adult photos

   Choose a clear picture of each person. Use your own image or get the other adult’s permission before uploading. Celebrity, public-figure, fictional-character and minor photos are not accepted. This two-person preset does not support pets.

2. Assign the two roles

   Put the person waiting on the dock in the first upload position and the singer in the second. Both photos are needed for the main sequence. If the roles are reversed, swap them in the editor before you submit.

3. Check the format and cost

   The main request is a 15-second, 4:3 MP4 without sound. Optionally see the free watermarked scene still without signing in. The still is not the video's first frame. Then sign in, confirm photo permission and review the credit cost for the 15-second video. One video is $4.99, or three are $9.90, as one-time purchases with optional monthly plans. Without signing in, see one free watermarked scene still from two photos. It is not the first frame of the video. Sign in and pay to make the separate 15-second video.

4. Generate and inspect

   Submit the request and check My creations for its status. Processing time can change with the queue. Download a completed file and watch it before sharing: AI can change the people, shot order, hands and cuts. A failed or blocked generation restores the credits used for that attempt.

5. Add sound when you post

   The two-photo MP4 is silent. In TikTok or Reels, add licensed audio your account is allowed to use, review the platform’s AI disclosure options, and publish there. The original song is not included here, and lip sync will not match its words or beat.

What to look for in your photos

A recognizable face is easier to assess when the source photo is sharp, well lit and unobstructed. Check that you selected the intended file for each role before paying for a generation. If a friend is involved, ask them about this specific use rather than assuming permission because the photo is already on your phone. The photos go to the generation provider; its own retention and training rules apply. The browser temporarily keeps the draft during sign-in and payment.

If you use the free preview, the two photos and the preview image are kept in private storage for 24 hours and then deleted. Otherwise this site does not keep input photos in its own database or private output storage. A completed video stays in My creations until you delete it. Those details matter if you are deciding whether to upload personal pictures. The privacy policy has the fuller processing explanation. Do not use a photo of a child or a public figure to test the scene.

• First position: the adult who waits on the dock.
• Second position: the adult who sings.
• Review both faces and the role order before you submit the paid job.

Know what the file contains

The two-photo result is a 15-second 4:3 file, not a 9:16 export. This site does not convert it. A vertical post letterboxes the frame unless you crop it and keep both people visible. It is silent: add licensed audio when posting. Lip sync will not match the original song.

The separate one-photo option on the homepage is a different scene: one adult singing on a dock in a 10-second vertical continuous shot. Its model may include audio. Do not use that option if your goal is the two-person dock-to-beach sequence. Neither option supplies the original song. This site does not add a watermark to your video.

If the request fails or the result differs

Check the job status in My creations before starting another attempt. A pending job is not the same as a failed one; queue time varies. When a job fails or the provider blocks it, the credits used for that attempt are restored. That credit restoration does not automatically refund a purchased pack. If the model refuses a photo, use another permitted adult image and keep the same subject rules.

A completed video may not preserve every facial detail or requested cut. The singer’s mouth is not driven by the original track, so do not expect word-for-word lip sync after adding music. Review the downloaded MP4 for people, framing, transitions and any sound you add in the posting app. Share only when the people pictured have agreed to this use.