Good starting point
- One person in each image, with permission to use the photo.
- A clear face looking toward the camera.
- Even lighting and visible natural facial detail.
- Head and upper body in frame, with the intended outfit visible.
Rain Dance AI Video Generator
Two photos. Your seaside duet. Reference audio included.Put your two people in the Rain Dance seaside duet. Upload one photo for each role and follow the reference's performance, camera work, and original audio.
Loading…
What is the Rain Dance AI trend?
Two people, a wooden pier, seaside close-ups, and a shared performance set to Raindance by Dave and Tems.
This template recasts the two characters in a prepared 28-second reference. Photo 1 replaces the man and Photo 2 replaces the woman, following their roles as the shots change.
Watch Dave and Tems' official videoThe preview shows the source reference. Your photos guide each character's face and outfit; the clip guides the movement and camera. Generated details can vary.
The song, the seaside scene, your version
The music supplies the mood. The two roles give you a clear place in the scene.
Raindance by Dave and Tems
Universal Music announced the official Raindance video on January 9, 2026, and identified Lagos, Nigeria as its filming location.
Read Universal Music's announcement- The reference
From the pier to the sea
The prepared clip moves between the pier, expressive close-ups, and both characters together. Watch it to see the performance your version follows.
Watch Dave and Tems' official video - Your version
Give each role a new face
Add a separate photo for each character. Their role stays tied to the photo even when the camera angle or screen position changes.
How to make a Rain Dance AI video
Choose the cast, check the settings, and review the finished duet.
Assign your two photos
Photo 1 replaces the man; Photo 2 replaces the woman. Use one clear person per photo and show the outfit you want the model to follow.
Choose length and quality
Start with 28 seconds at 720P, or choose 5 or 15 seconds and 480P. The video uses 4:3 framing. Check the live total credit quote before generating.
Watch and download
Watch the whole video with sound. Check both faces during turns, the outfits, transitions, and audio timing before downloading the version you want to keep.
Choose photos for both roles
Each photo supplies one character's identity and clothing. The two photos do not need to be taken together.
Avoid these
- Group photos with several competing faces.
- Sunglasses, masks, or hands covering the face.
- Dark, blurry, tiny, or heavily filtered portraits.
- Tight crops that cut off the head or hide the outfit.
Your cast in the seaside duet
A prepared scene brings both roles, the performance, and the reference audio into one workflow.
Two defined roles
Each photo follows one character through the shots. Assign by role, even when the performers change position.
Your faces and outfits
Each uploaded image guides its character's appearance and clothing. Choose photos that show the look you want to carry into the scene.
A choice of lengths
Choose 5, 15, or 28 seconds at 480P or 720P. The 4:3 frame follows the reference composition.
Reference audio included
The finished download includes the reference video's audio for the selected duration, with no separate audio-editing step.
Check the total before you generate
Length and quality determine your credit quote. The tool shows the total, including automatic processing, before you start.
5 or 15 seconds · a shorter version
Select a shorter length and either 480P or 720P. Review the current quote and use your balance or a credit pack.
28 seconds · the full reference length
The default is 28 seconds at 720P. Submit one task and receive one complete video with reference audio. Check the live total for your chosen quality.
The provider documentation describes the model. Your total credit charge is the live quote shown in this tool. Wan3 Prime model documentation
Rain Dance AI vs CapCut templates vs DIY prompts
Compare a prepared two-role reference with an editing template or your own AI prompt workflow.
Rain Dance AI
- What you provide
- Two character photos. The reference and prompt are prepared.
- Scene and movement
- The seaside reference guides both roles, movement, and camera. Results can vary.
- Output choices
- 5, 15, or 28 seconds; 4:3; 480P or 720P; reference audio included.
- Cost planning
- Review the live credit total for your settings before generating.
CapCut templates
- What you provide
- Add the photos or clips requested by your chosen template.
- Scene and movement
- Uses the selected template's effects, transitions, and animations.
- Output choices
- Check the template's length and your editor's export options.
- Cost planning
- Free and paid templates. Check the selected template's price and requirements.
DIY AI prompts
- What you provide
- A prompt, subject photos, and any references your model supports.
- Scene and movement
- Describe the scene and actions; tune the prompt and references yourself.
- Output choices
- Length and resolution depend on the model you select.
- Cost planning
- Check generation and retry charges with your provider.
Explore another duet or prepare your photos
Try a different two-person scene or review the photo checklist before casting your next video.
Alors on Danse AI Video
Cast a keyboard player and a partner in the white-studio scene.
Explore generatorHotel Lobby AI
Put a different pair into the orange-booth rap performance.
Explore generatorChoose better reference photos
Check face detail, light, framing, and anything that might hide the subject.
Explore generatorRain Dance AI video questions
What is the Rain Dance AI video trend?
It is a two-person seaside remake associated with Raindance by Dave featuring Tems. This template follows a prepared 28-second reference with pier shots, close-ups, and both characters together. The preview is the source reference, not a generated result.
Which role does each photo replace?
Photo 1 replaces the man in the reference; Photo 2 replaces the woman. The assignment follows the character across shots, not a fixed left or right side of the screen. Use a separate image for each role.
Which model is used, and do I need a prompt?
The template uses Wan3 Prime reference-to-video through fal. The prompt and reference clip are prepared, so you only need to add the two photos and choose your settings.
How long is the video, and what quality can I choose?
Choose 5, 15, or 28 seconds, with 480P or 720P output in 4:3. The default is 28 seconds at 720P. The full-length option is handled as one task and delivered as one complete video.
How much does a Rain Dance video cost?
Choose your length and quality to see the current total credit quote, including automatic processing. Use your existing balance or a credit pack. Check the quote again before starting another attempt.
Does the video include the original reference audio?
Yes. The download includes the source reference's audio for your selected duration. Play the result to check timing. Generating a video does not grant additional rights to the music or reference footage.
Will faces, outfits, and movements match exactly?
The photos guide each character's face and outfit; the reference guides actions, expressions, camera moves, and shot changes. This is generative video, so exact matching is not guaranteed. Check faces during turns, clothing, transitions, and audio timing across the full result.
What photos should I use?
Use one person per photo, with a sharp face, even light, and the intended outfit visible. Avoid group shots, strong filters, and objects covering the face. Use your own photos or images you have permission to use, including the person's permission when casting someone else.