Photos or audio · Wan 3.0 · Seedance 2.5 · 9:16
AI Music Video Generator
Turn two photos — or your own audio — into a vertical music video in minutes. Wan 3.0 or Seedance 2.5, 9:16 for TikTok and Reels. Failed renders refund credits.
Make my AI music videoYour two faces
A single face per picture. Left leads, right backs up.
Tip: Drag photos into the boxes, or press Ctrl+V (⌘V) to paste.
Set
Take and sound
Price: 6 credits. Done in roughly 5 to 15 min; if it fails, the credits come back. See prices
How it works
How to make an AI music video in 3 steps
Making an AI music video takes no filming and no editing. The Create button shows the credit price before you spend it.
- 1
Add two photos, or a photo and your audio
Upload one clear, front-facing photo per person, or one shared shot of both. JPG, PNG and WEBP up to 20 MB work. For your own sound, add an MP3 or WAV up to 50 MB and pick the 2 to 15 second section to perform. Without audio, Seedance 2.5 writes the lyrics from a topic.
- 2
Pick a scene, a model and a shape
Choose Hotel Lobby, Luxury Lobby, Studio Booth or Street Cypher. Use Wan 3.0 for your own audio, or Seedance 2.5 for AI-written lyrics on any scene. Set 9:16 for TikTok and Reels, then pick 480p, 720p or 1080p and a length.
- 3
Create, download, post
Press Create and sign in, because renders use credits. Seedance usually finishes in 2 to 5 minutes and Wan 3.0 in 5 to 15. Download the MP4 and post it. If the render fails, the credits come back automatically.
Two inputs
AI music video from photos or audio: which input to start from?
Both paths start with photos. The question is where the sound comes from.
Searching for a music video from a photo and for one from audio sounds like two tools. Here it is one tool with two ways to supply the sound. The picture always comes from your photos. The sound is either an AI-written verse or a track you upload.
Start from photos alone when you have no song. Give the studio a topic of 3 to 120 characters, such as a birthday or a roast, and Seedance 2.5 writes the lines and performs them. Start from audio when the words already exist, like a verse you wrote or a voice note. That path runs on Wan 3.0 and Hotel Lobby, and you pick a 2 to 15 second section.
| Photos only | Your own audio | |
|---|---|---|
| Where the sound comes from | AI-written lyrics from your topic | Your MP3 or WAV |
| Model | Seedance 2.5 (also Mini, Fast, 2.0) | Wan 3.0 |
| Scenes | All four | Hotel Lobby |
| Clip length | 5 to 30 seconds on Seedance 2.5 | 2 to 15 seconds, the section you pick |
| Best for | Birthdays, gifts, roasts, quick posts | Your own song, a hook, a verse you wrote |
| Typical wait | 2 to 5 minutes | 5 to 15 minutes |
Free, exactly
How free is the AI music video generator, and where does paying start?
Free here means the studio and the price check. A render uses credits.
Opening the studio, uploading photos or audio, choosing a scene and reading the exact credit price are all free, with no card. You sign in only when you press Create, because a render spends credits. The button shows the total before anything is charged.
Rendering is the part that costs. Credits are sold in packs from $6.99, and a 12-second 480p draft is the cheapest test: 6 credits on Wan 3.0 and 17 on Seedance 2.5. Trial credits for new accounts depend on the current offer, so check the Create button after you sign in.
Paid renders are HD with no watermark, and you can download and post them anywhere. If a render fails, the credits are returned.
| Step | Cost | Account needed |
|---|---|---|
| Open the studio, pick scene and model | Free | No |
| Upload photos or audio and trim a section | Free | No |
| See the credit price | Free | No |
| Render an AI music video | Credits | Yes, sign in at Create |
| Download an MP4 with no watermark | Included with paid credits | Yes |
| A render that fails | Credits returned | Yes |
Two models
Wan 3.0 vs Seedance 2.5 for an AI music video
Each model unlocks different options, so the choice depends on what you bring.
Wan 3.0 runs on Hotel Lobby only. In return it performs your own audio, a 2 to 15 second section of an MP3 or WAV, and can copy the movement of a short clip you upload. It does not write lyrics. For a 12-second video it costs 6 credits at 480p, 10 at 720p and 20 at 1080p.
Seedance 2.5 runs on all four scenes and writes the lyrics from a topic. It reaches 1080p, up to 30 seconds and a 21:9 shape that Wan 3.0 lacks. For 12 seconds it costs 17 credits at 480p, 37 at 720p and 93 at 1080p. Seedance Mini, Fast and 2.0 cost less at every resolution.
The studio estimates 2 to 5 minutes for Seedance and 5 to 15 for Wan 3.0, because Wan queues longer. We have not published a failure rate; failed renders are refunded. Use Wan 3.0 when the sound is yours and Hotel Lobby fits, and Seedance 2.5 for any other scene, AI lyrics, long clips or a 1080p final.
No app to install
AI music video maker online vs a phone app
The studio runs in your browser on a phone or a laptop.
There is nothing to download. Open this page in Safari, Chrome or any current browser, upload from your camera roll or files, and the render runs on our servers rather than on your device. Your phone does not heat up, and an old handset works as well as a new one.
A phone app sits on your home screen and can open the camera directly, but it needs storage, updates and an app store account. The web studio trades that for zero setup. Audio trimming happens in the browser before upload, and when the render finishes you download the 9:16 MP4 and share it to TikTok, Reels or Shorts.
Other ways to do it
AI music video vs face swap, song generators, editing it yourself and hiring someone
Five ways to get a music video with you in it. They solve different problems, and an AI music video is the fastest.
Face swap
A face swap needs a source video. You find a clip of someone performing and paste your face onto it. The set, the moves and the body belong to somebody else. Here a video model generates a new clip from your photos and a scene, so there is no source clip to find, license or credit.
Song-only generators
An AI song generator gives you audio and nothing to look at. You still need footage, and matching a mouth to a track is your job. In this studio the voice and the mouth movement come out of the same render, and with Wan 3.0 the voice can be your own recording.
Editing it yourself
Cutting stock footage to a song in a video editor gives full control and takes hours, and you need footage that matches the song. An AI music video render takes minutes, but you control less: scene, model, length and shape, not every frame.
Hiring a video editor or a freelancer
A person can storyboard, shoot and cut, and the result can be excellent. It is also slow and priced per project, which is hard to justify for a birthday clip or a social post. Hire someone for a launch asset. Use an AI music video when you have photos or a recording and want a vertical clip today.
With your face
AI music video with your face: which photos work
The video follows the faces in your photos, so the photo matters more than any setting.
The model is told to keep faces and outfits matching your photos, so what you upload is what appears. Use one clear, front-facing photo per person: eyes visible, face fully in frame, even light, no sunglasses and no hand across the mouth. A sharp phone photo beats a heavily filtered one.
We have no list of face shapes that work better. What matters is visibility, not shape. Large, front-lit faces looking roughly at the camera give the steadiest results. Strong shadows, extreme angles and tiny faces in a wide shot are the usual causes of drifting features.
With one photo of both people, upload it as a shared photo: the person on the left leads and the right backs up. Crop a group shot to two people first, and only upload photos of people who agree to appear.
Examples
Two AI music video examples and their specs
What went in and what came out, so you can judge the result yourself.
Friends on one stage
Input: two portrait photos, Hotel Lobby scene. Output: 12 seconds, 9:16, 480 by 854 pixels.
A couple, one shared photo
Input: a single photo with both people, Hotel Lobby scene. Output: 12 seconds, 9:16, 480 by 854 pixels.
Both are 480p, 12-second sample renders from our gallery, which cost 6 credits on Wan 3.0 and 17 on Seedance 2.5. We have not recorded how many attempts each took, so we will not invent that number. Play them with sound on and watch whether the faces stay stable when the heads turn. Test your own AI music video at 480p first: if the faces hold there, they hold at 1080p, for far fewer credits.
Price
AI music video price and free trial
Credits pay for renders. You see the price before you create.
Credit packs start at $6.99 for 10 credits, enough for a 12-second 720p video on Wan 3.0. Monthly plans start at $29.90 for 100 credits, and packs never expire. Paid renders are HD with no watermark. Credits for a failed render come back automatically.
FAQ
AI music video generator FAQ
Price, app, audio, length, format and refunds.
Is the AI music video generator free?
Opening the studio, uploading photos or audio and seeing the exact credit price are free. Rendering an AI music video uses credits, shown on the Create button before you start. Credit packs start at $6.99 for 10 credits. Paid renders have no watermark, and credits from a failed render are returned.
Do I need to install an app to make an AI music video?
No. The generator runs in your browser on a phone or a computer, and the video renders on our servers, so nothing is installed and your device does not slow down. Upload from your camera roll or files, then download the finished MP4 and share it to TikTok, Reels or Shorts.
Can I make an AI music video from my own MP3?
Yes. Choose Wan 3.0 on the Hotel Lobby scene, upload an MP3 or WAV up to 50 MB and pick the section you want. The lead voice performs it, and the section is trimmed in your browser before upload. With no audio, Seedance 2.5 can write lyrics from a topic instead.
Can the AI music video use a full-length song?
You can upload a full song file, but each render performs one section of 2 to 15 seconds that you choose, such as the hook or a verse. It does not render a whole three-minute song in one go. To cover more of a song, render several sections and join the clips in an editor.
How long can an AI music video be?
Lengths run from 5 to 15 seconds on the models, and Seedance 2.5 adds 20 and 30 seconds. With your own audio on Wan 3.0 the performed section is 2 to 15 seconds. The credit price rises with length and resolution, and the Create button shows it before you spend anything.
What resolution and aspect ratio does the AI music video have?
Choose 480p, 720p or 1080p. Shapes are 9:16, 16:9, 1:1, 3:4 and 4:3, plus 21:9 on Seedance. Pick 9:16 for TikTok, Reels and Shorts. At 480p a 9:16 video measures 480 by 854 pixels. A higher resolution costs more credits, so test at 480p first.
How many photos do I need for an AI music video?
Use either two photos, one per person, or one shared photo that shows both people. The left person leads and the right person backs up. The studio is built around two performers, so crop a group shot down to two. Each person should have one clear, front-facing face.
Is this AI music video generator a face swap?
No. A face swap pastes your face onto an existing video. Here a video model generates a new clip from your photos and a scene description, and it is told to keep faces and outfits matching the photos. There is no source clip behind it. The Move clip option only copies body movement.
What happens if the render fails?
Credits for a failed generation are returned to your balance automatically, so a failed render costs nothing. Temporary provider errors are retried first, and the refund happens once the task is finally marked as failed. You can then change the photos or settings and create again.
Can I use an AI music video commercially?
Yes. Videos have no watermark and you can download and post them anywhere, including social accounts and ads. You are responsible for having the right to use every photo and audio file you upload, and for getting consent from the people in them. Do not upload music you do not have rights to.
Updates
Published and updated
Published October 30, 2026. Last updated October 30, 2026.
- Page published. Credit prices and model limits match the studio on this date.
Turn two photos into a music video
Pick a scene, choose a model, and see the credit price before you create.