Intro video · How to Make an AI Music Video: Your Song, AI Lip Sync and a 6-Part Prompt Formula (1 min 25 sec)

AI Music Video Prompts for singing shots, lyric videos and visualizers

Every example on this page is a real video we generated, shown with the exact prompt, model, settings and credits. Every performer is AI-generated. Seedance 2.5 and Wan 3.0 make their own music and voices from the prompt; to have a singer follow a song you already have, use AI Lip Sync, as in the song-first example below.

Intro video
1 min 25 sec
Example models
Seedance 2.5 · Wan 3.0 · AI Lip Sync
Aspect ratio
16:9
Prompts
10
Total example credits
1,784

All examples were made in Bandicam AI Studio: the rooftop band video and the 15-second synthwave shot list with Seedance 2.5, the singing shot and the 5-second scenes with Wan 3.0, and the song-first singer with AI Music Generator, GPT Image 2 and AI Lip Sync. The performers are fictional and AI-generated, and no real artists or songs are used.

How to write an AI music video prompt

Fill in these six parts in order. Each video model makes its own soundtrack, so describe the music you want to hear, quote any line to be sung, and write “No vocals” for instrumental shots. For one 5-second shot, keep one subject, one action and one camera move.

  1. Music and mood

    Name the genre, tempo and mood, and say whether anyone sings.

    for an upbeat indie-pop song … upbeat indie-pop with drums, bass and bright guitar
  2. Performer or visual subject

    A fictional adult performer, a band, an animal, a place or an abstract shape. Describe looks and clothes, never a real artist.

    a woman in her late 20s with shoulder-length dark wavy hair and a shimmering silver jacket
  3. Setting, light and look

    Where it happens, the time of day, the main light and the visual style.

    a city rooftop at dusk strung with warm paper lanterns
  4. Shot and camera

    One camera move per 5-second shot. For longer clips, write timed ranges with one shot each. Cut to your song later in an editor.

    Shot 1 (0-3s): slow crane shot rising over a city rooftop
  5. Sound and lyrics

    Quote one short line to be sung, or say there is no singing.

    from 6s the singer's voice, heard over the band, sings: “Every lantern's swaying to the drum tonight.”
  6. Exclusions

    List what to leave out: text, logos and signs. For instrumental shots, add “No vocals, no singing, no speech.”

    No subtitles, no on-screen text, no logos, no watermark
  • Each clip gets its own sound

    Text-to-video clips generate their own audio. Two singing clips of the same singer won't share one song or voice, so don't plan to cut them together as one performance.

  • Use AI Lip Sync for your own song

    To show a singer performing a song you already have, give AI Lip Sync one front-facing photo with the mouth clearly visible and up to 30 seconds of audio. The face moves along with that audio.

  • Keep words out of the picture

    Lyrics, titles and signs often come out garbled. Ask for “no on-screen text, no signs,” then add lyrics as captions in your editor.

  • Describe gear and places, not brands

    Say “plain instruments with no logos” and describe shapes and colors. Famous cars and skylines can still show up, so check every result: our first synthwave, dance and singer takes had to be redone.

  • Cut to the beat in editing

    Words like “in time with the music” set the mood, but the clips are not synced to your track. Line up cuts and pulses with the music in an editor.

Change one thing

We took one example from this page, changed a single part and generated it again. Everything else in the prompt, the model and the settings stayed the same. The highlighted words are what changed. When a result is not what you want, change one thing at a time like this.

Changed · music style and vocal delivery

Same singer, slow ballad

Original · Rooftop singer: a 5-second singing shot
Result with only the music style and vocal delivery changed
  • takes a quick breath, then sings with full energy, her jaw opening wide on the long vowels and her brows lifting on the high notesings softly with her eyes half closed, holding the long notes
  • Upbeat pop backing track with punchy drums, bright synths and bassSlow piano ballad backing with soft strings

The prompt keeps the rooftop setting, camera and quoted line, but asks for a piano ballad and a softer delivery: her eyes stay mostly closed. Her face moves less than in the upbeat version, and the lanterns and trees came out a little different. This is our second try: the first, on a city rooftop, showed a famous real skyline.

Changed prompt · 691 characters
Music video performance shot, 16:9, one continuous shot. A young adult woman pop singer in her late twenties, with shoulder-length dark wavy hair and a shimmering silver jacket over a black top, stands at a vintage microphone in a rooftop garden at dusk, with strings of warm paper lanterns, potted trees and a soft violet sunset sky behind her, no buildings or city skyline. Medium close-up with a slow push-in. She sings softly with her eyes half closed, holding the long notes, singing clearly in English: "Turn it up, let the night run wide." Slow piano ballad backing with soft strings, her solo female lead vocal in front. No other voices, no subtitles, no captions, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Credits
48

November 21, 2026 · Wan 3.0

Changed · music genre

Same shot, now rapping: an AI rapper video

Original · Rooftop singer: a 5-second singing shot
Result with only the music genre changed
  • A young adult woman pop singerA young adult woman rapper
  • takes a quick breath, then sings with full energy, her jaw opening wide on the long vowels and her brows lifting on the high note, singing clearly in English: "Turn it up, let the night run wide."raps one quick line straight to the camera, nodding on every beat and making small chest-level gestures with her free hand, her mouth shaping each word crisply, rapping clearly in English: "Paper lanterns swaying while I'm riding every beat tonight."
  • Upbeat pop backing track with punchy drums, bright synths and bass, her solo female lead vocal in front.Hip-hop beat with deep bass, crisp hi-hats and a hard snare, her solo female rap vocal in front.

Three swaps turn the pop singer into a rapper: the word “rapper,” a rapping action with a new line we wrote, and a hip-hop beat. The rooftop garden, microphone and push-in stayed, and she nods and gestures with her free hand, but Wan 3.0 drew a different woman: a text prompt doesn't keep a face from one video to the next. To keep the same performer in every shot, start from one photo with AI Lip Sync, as in the song-first example.

Changed prompt · 823 characters
Music video performance shot, 16:9, one continuous shot. A young adult woman rapper in her late twenties, with shoulder-length dark wavy hair and a shimmering silver jacket over a black top, stands at a vintage microphone in a rooftop garden at dusk, with strings of warm paper lanterns, potted trees and a soft violet sunset sky behind her, no buildings or city skyline. Medium close-up with a slow push-in. She raps one quick line straight to the camera, nodding on every beat and making small chest-level gestures with her free hand, her mouth shaping each word crisply, rapping clearly in English: "Paper lanterns swaying while I'm riding every beat tonight." Hip-hop beat with deep bass, crisp hi-hats and a hard snare, her solo female rap vocal in front. No other voices, no subtitles, no captions, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

Changed · weather and time

Same room on a snowy morning

Original · Lofi room: a calm background clip
Result with only the weather and time changed
  • at night, rain running down the window, city lights blurred outsideon a snowy morning, snow drifting past the window, pale daylight outside
  • gentle rain, steam and record motiongentle snowfall, steam and record motion
  • soft vinyl crackle and quiet rainsoft vinyl crackle and a quiet winter hush

Only the weather and time of day changed in the prompt. The plant, mug and turntable are in similar places, but the window, lamp and bedding differ, so the two clips won't join as one loop.

Changed prompt · 553 characters
16:9 lofi music video background clip, one continuous shot, locked-off camera: a cozy empty bedroom on a snowy morning, snow drifting past the window, pale daylight outside, a desk lamp glowing warm, a small plant, a mug with rising steam and a record with a plain black label spinning slowly on a turntable. The camera and furniture stay still and the lighting stays steady, with only gentle snowfall, steam and record motion. Mellow lofi hip-hop beat with soft vinyl crackle and a quiet winter hush. No people, no vocals, no speech, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

Music video terms at a glance

You can paste the wording from the middle column straight into a prompt. Click a term to jump to the example that uses it.

TermPrompt wordingWhat it does on screen
Performance shota singer at a vintage microphoneThe artist sings or plays on camera.
Cutaway (B-roll)one slow-motion close-up of a drumstick striking a cymbalA detail shot to cut to between performance shots.
Timed shot list0-3s: … 3-6s: …One shot per time range, in the written order.
Lip syncAI Lip Sync: one photo + your audioThe mouth follows a song you already have.
Locked-off shotthe camera stays stillThe frame stays fixed, so all the movement comes from the subject.
Loop clipThe camera and furniture stay stillOnly small things move, so the clip is easy to repeat with a crossfade.
Lyric video backgroundthe upper half is plain open skyA calm, empty area where lyric captions can sit.
Visualizerrings pulse outward from the centerAbstract shapes that move like music, for a visualizer background.

AI music video prompts: 10 examples

Click a thumbnail to play the example video. Every prompt is shown exactly as it was used to make the video, so you can copy it or send it straight to Bandicam AI Studio.

The colors show which part of the prompt formula above each phrase belongs to.

  • 1Music and mood
  • 2Performer or visual subject
  • 3Setting, light and look
  • 4Shot and camera
  • 5Sound and lyrics
  • 6Exclusions
  • CutTiming and cuts

01 · Three shots · band reveal · singer from behind

Rooftop band: a 10-second music video in three shots

A crane shot rises over the parapet to reveal the band under paper lanterns, cuts to a slow-motion cymbal, then shows the singer from behind as she raises her hand. Filming the singer from behind means her mouth never has to match the words, the hardest part of a singing shot. The prompt asks Seedance 2.5 to make the band music and a sung line itself.

Prompt · 1011 characters
16:9 music video, 10 seconds, three shots, for an upbeat indie-pop song, warm dusk look, light film grain.
Shot 1 (0-3s): slow crane shot rising over a city rooftop at dusk strung with warm paper lanterns, revealing a small band playing in the middle: a drummer, a guitarist, a bass player and a singer at a vintage microphone, all with plain instruments and no logos, a skyline of unnamed glass towers glowing beyond, no famous landmarks.
Shot 2 (3-6s): one slow-motion close-up of a drumstick striking a cymbal, warm lantern reflections shimmering across the metal.
Shot 3 (6-10s): medium-wide shot from behind the singer, a woman in her late 20s with shoulder-length dark wavy hair and a shimmering silver jacket, as she raises one hand toward the violet sky and the band hits the chorus.
Sound: upbeat indie-pop with drums, bass and bright guitar; from 6s the singer's voice, heard over the band, sings: "Every lantern's swaying to the drum tonight."
No subtitles, no on-screen text, no logos, no watermark.
Model
Seedance 2.5
Aspect ratio
16:9
Length
10 sec
Resolution
720p
Sound
On
Credits
740

November 21, 2026 · Seedance 2.5 · first result, published as generated

02 · Five timed shots · synthwave · no vehicles

Synthwave night: a 15-second shot list

Five 3-second ranges go from a rainy neon street to a tunnel of light tubes, a woman turning in a brick alley, raindrops on glass and a retro sunset over a glowing grid, in the written order. The prompt says “no cars, no vehicles”: with a car in it, our first try came out looking like a famous real model.

Prompt · 1235 characters
[Style] 15-second retro synthwave music video, 16:9, neon magenta and teal color grade, light film grain, 1980s feel.
[Must exclude] No cars, no vehicles, no vocals, no singing, no humming, no speech, no subtitles, no on-screen text, no signs, no graffiti, no title cards, no logos, no watermark.
[Shot list]
0-3s: slow low push along an empty rain-soaked city street at night, magenta and teal neon light reflecting in puddles on the wet asphalt, dark buildings with no signs or lettering.
3-6s: a long dark tunnel where glowing magenta and teal light tubes switch on one after another, rushing toward the camera.
6-9s: a woman in her 20s in a cropped silver jacket stands in a narrow alley between plain wet brick walls lit by neon, and turns slowly to face the camera as the beat drops; she does not sing or speak.
9-12s: close-up of raindrops sliding down a window, neon lights behind the glass blurred into soft pulsing bokeh.
12-15s: the camera flies low over a glowing grid-lined plain toward a huge striped retro sunset, then rises and pulls back wide.
[Sound] Driving synthwave instrumental at about 110 BPM with gated drums, pulsing bass and a bright lead synth, the beat dropping around 6s. No vocals, no singing, no speech.
Model
Seedance 2.5
Aspect ratio
16:9
Length
15 sec
Resolution
480p
Sound
On
Credits
494

November 21, 2026 · Seedance 2.5 · second result: the first, with a car, looked like a famous real model · blurred neon signs with letter-like marks still appear in the first 3 seconds despite “no signs”

03 · Generated vocal · medium close-up · push-in

Rooftop singer: a 5-second singing shot

Wan 3.0 makes the voice and the mouth movement in the same take, so they belong together. It writes its own melody, so use a shot like this where its own sound plays, not under a different song. Words like “jaw opening wide” gave a fuller singing face than the soft ballad version below.

Prompt · 781 characters
Music video performance shot, 16:9, one continuous shot. A young adult woman pop singer in her late twenties, with shoulder-length dark wavy hair and a shimmering silver jacket over a black top, stands at a vintage microphone in a rooftop garden at dusk, with strings of warm paper lanterns, potted trees and a soft violet sunset sky behind her, no buildings or city skyline. Medium close-up with a slow push-in. She takes a quick breath, then sings with full energy, her jaw opening wide on the long vowels and her brows lifting on the high note, singing clearly in English: "Turn it up, let the night run wide." Upbeat pop backing track with punchy drums, bright synths and bass, her solo female lead vocal in front. No other voices, no subtitles, no captions, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · second result: the first, on a city rooftop, showed a famous real skyline, so the city was left out · the earlier city-rooftop prompt on Seedance 2.5 looked overacted

Starting photo (AI-generated, GPT Image 2, 51 credits)
Starting photo (AI-generated, GPT Image 2, 51 credits)

04 · AI Music Generator · photo · AI Lip Sync

Song first: a singer who follows your own track

We made a 35-second song in AI Music Generator (Quick mode, Pop, Uplifting) from the lyrics shown, took the 15 seconds from 0:14 to 0:29, the start of the chorus, and gave AI Lip Sync that audio with the AI-generated singer photo. The video keeps the supplied song section, and her mouth, head and hands move as she sings. The prompt box holds the song lyrics; AI Lip Sync itself takes no prompt.

Prompt · 230 characters
[Verse]
Paper lanterns on the rooftop line
The city hums beneath a violet sky
[Chorus]
Turn it up, let the night run wide
We are glowing for a little while
Turn it up, every window bright
We keep glowing, glowing through the night
Model
AI Lip Sync
Aspect ratio
16:9
Length
15 sec
Resolution
720p
Sound
On
Credits
214

November 21, 2026 · AI Music Generator + GPT Image 2 + AI Lip Sync · same song and section, second photo: the first, with a city behind her, showed a famous real skyline

Try AI Lip Sync

05 · Three players · dolly in · garage rock

Garage band: a performance shot

The camera starts wide on the drummer, bassist and guitarist and dollies in to the guitarist's hands. Asking for plain amps, a plain kick drum and an “invented” guitar kept logos away, but the guitar still came out with a familiar offset body shape, so check the gear in every result.

Prompt · 647 characters
16:9 music video performance shot, one continuous shot, no cuts: an indie rock band of three adults plays in a cluttered garage at night lit by one bare bulb and string lights: a drummer in a gray hoodie behind a plain black kick drum, a bass player and a guitarist with an invented asymmetrical cream electric guitar, its headstock in shadow, in front of plain black amplifiers with bare grilles. Slow dolly from the drummer's cymbals toward the guitarist's hands on the strings, dust floating in the light. Loud garage rock instrumental with crashing drums, distorted guitar and driving bass. No vocals, no singing, no speech, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

06 · Locked-off · full body · keeps moving

Dance cut: a full-body shot in a warehouse

A still camera and “keeps dancing the whole time” gave steady movement for all 5 seconds. Our two rooftop versions barely moved and showed a famous skyline, so this one uses an empty warehouse with colored side lights. For choreography and copying moves from a video, see the dance video prompts page.

Prompt · 497 characters
16:9 music video dance shot, one continuous shot, no cuts: a young adult man in a loose black jacket dances in an empty warehouse with a concrete floor and light haze, lit by magenta and blue colored lights from both sides, no windows. He keeps dancing the whole time with a bouncy side-to-side two-step, shoulders rocking and arms swinging wide. Full-body wide shot, the camera stays still. Hard-hitting hip-hop instrumental beat with deep bass. No vocals, no speech, no signs, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · third result: two rooftop versions barely moved and showed a famous skyline, so the setting changed

07 · Locked-off · rain · no people

Lofi room: a calm background clip

Only the rain, steam and record move while the camera stays fixed, which makes the clip easy to loop with a crossfade in an editor. The room is empty, so it doesn't lean on any well-known lofi character. Loop or replace the soundtrack separately in your editor.

Prompt · 535 characters
16:9 lofi music video background clip, one continuous shot, locked-off camera: a cozy empty bedroom at night, rain running down the window, city lights blurred outside, a desk lamp glowing warm, a small plant, a mug with rising steam and a record with a plain black label spinning slowly on a turntable. The camera and furniture stay still and the lighting stays steady, with only gentle rain, steam and record motion. Mellow lofi hip-hop beat with soft vinyl crackle and quiet rain. No people, no vocals, no speech, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

08 · Drone glide · open sky · no text

Lake at golden hour: a lyric video background

The prompt never says “lyric video,” which can invite drawn words; it asks for open sky and “no on-screen text” instead. In this result the horizon sits near the middle rather than low, but the sky stays plain, with room for captions.

Prompt · 355 characters
16:9 music video background, one continuous shot, no cuts: a slow drone glide just above a calm lake at golden hour, low mist on the water, pine hills on both sides, the horizon low in the frame so the upper half is plain open sky. Gentle acoustic guitar and piano instrumental. No vocals, no singing, no speech, no on-screen text, no subtitles, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

09 · Abstract · centered · camera still

Chrome rings: a visualizer background

Magenta and cyan rings ripple out from the center on black. They pulse at a steady pace, not in sync with any track; to make the pulses hit your beat, cut or speed-ramp the clip in an editor.

Prompt · 388 characters
16:9 abstract music visualizer background, one continuous shot, centered, the camera stays still: glossy liquid-chrome ripples and glowing magenta and cyan rings pulse outward from the center at a steady, dance-like pace on a pure black background. Punchy electronic dance instrumental with a steady four-on-the-floor kick and bright synth chords. No vocals, no speech, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

10 · 2D animation · tracking shot · folk mood

Fox in a glowing forest: an animated music video shot

Describing the look (“hand-painted 2D” and “gentle storybook feel”) instead of naming a studio keeps the style your own. The result reads more like a clean storybook illustration than watercolor, and the fox keeps the same shape and colors for all 5 seconds.

Prompt · 403 characters
16:9 hand-painted 2D animated music video shot, one continuous shot: a small red fox carrying a tiny lantern walks along a mossy path through a night forest full of glowing blue mushrooms and fireflies, the camera tracking beside it. Soft watercolor textures, gentle storybook feel. Warm acoustic folk instrumental with fingerpicked guitar and soft glockenspiel. No vocals, no speech, no text, no logos.
Model
Wan 3.0
Aspect ratio
16:9
Length
5 sec
Resolution
480p
Sound
On
Credits
48

November 21, 2026 · Wan 3.0 · first result, published as generated

Model and settings for each prompt

The values we chose when we made each example video. Credits show the charge for one generation of that video, or a quoted price where a note below the table says so.

PromptModelAspect ratioLengthResolutionSoundCredits
01Rooftop band: a 10-second music video in three shotsSeedance 2.516:910 sec720pOn740
02Synthwave night: a 15-second shot listSeedance 2.516:915 sec480pOn494
03Rooftop singer: a 5-second singing shotWan 3.016:95 sec480pOn48
04Song first: a singer who follows your own trackAI Lip Sync16:915 sec720pOn214
05Garage band: a performance shotWan 3.016:95 sec480pOn48
06Dance cut: a full-body shot in a warehouseWan 3.016:95 sec480pOn48
07Lofi room: a calm background clipWan 3.016:95 sec480pOn48
08Lake at golden hour: a lyric video backgroundWan 3.016:95 sec480pOn48
09Chrome rings: a visualizer backgroundWan 3.016:95 sec480pOn48
10Fox in a glowing forest: an animated music video shotWan 3.016:95 sec480pOn48
Total1,784

Song-first singer: 214 credits in total = song 12 (AI Music Generator, Quick) + singer photo 51 (GPT Image 2) + lip sync 151 (15 seconds, 720p).

Troubleshooting

Problems we ran into while making the examples on this page, with fixes from our tests and the model makers' official guides. The last column gives the source.

ProblemFixSource
The singer's mouth doesn't match my songText-to-video makes its own voice. Use AI Lip Sync with your audio and a front-facing photo where the mouth is clearly visible.Examples on this page
A famous car, guitar or skyline shows upLeave the object out, or describe it as plain and fictional, and check every result. We removed the car from the synthwave shot list, moved the dance shot indoors and put the singer in a rooftop garden.Examples on this page
Instrumental shots have singing or talkingWrite “No vocals, no singing, no speech.” For Seedance 2.5, say it in both the exclusions and the sound line.Examples on this page
Lyrics or signs come out as garbled textAsk for “no on-screen text, no signs” and add the lyrics as captions in editing. Blurry letter-like marks can still appear, as in the first seconds of the synthwave clip.Examples on this page
The dancer barely movesKeep the camera still and say the dancer keeps moving the whole time, with one simple step.Examples on this page

How to make an AI music video

  1. Start with the song

    Make a song from your own lyrics in AI Music Generator, or use a recording you have the rights to. Listen to it and pick the section you want to show.

  2. Make the shots

    For a singer who follows your song, use AI Lip Sync with a front-facing photo and that section. For band, dance, lofi, lyric video and visualizer shots, pick an example below, click “Use this prompt” and change the subject. These clips come with their own generated sound.

  3. Edit to the beat

    In a video editor, put your song under everything, mute or lower the generated sound, cut the shots to the beat and add the lyrics as captions. Check every mouth movement and caption before you export.

Try it yourself

The prompt from the “Rooftop band: a 10-second music video in three shots” example is already filled in. Edit it or write your own, then click “Generate video” to continue in Bandicam AI Studio.

0 / 2,000

AI music video prompts: frequently asked questions

Can AI make a music video?

Yes, one shot at a time. Seedance 2.5 and Wan 3.0 in Bandicam AI Studio make clips with sound, up to 30 seconds each; the rooftop band video here is one 10-second clip. For a whole song, make several clips and join them under your track in an editor.

Can an AI singer sing my song?

Use AI Lip Sync: upload one clear, front-facing photo and up to 30 seconds of your audio, and the face moves along with it. The song-first example here used 15 seconds of a song made in AI Music Generator. Text-to-video models such as Wan 3.0 make up a new tune and voice instead.

How do I make an AI lyric video?

Generate calm background shots with “no on-screen text,” like the lake example, then add the lyrics as captions in your editor. Lyrics drawn by the AI usually come out misspelled.

Will the video follow the beat of my music?

Not by itself. A prompt can describe a pace, but the clips are not synced to your track. Trim and cut them to the beat in an editor.

Can I make the music with AI too?

Yes. AI Music Generator in Bandicam AI Studio makes a song with vocals from your lyrics, or an instrumental track. Write your own lyrics and describe the style in your own words instead of naming an artist.

Can I use a real singer, band or song?

Don't name or imitate real artists, and don't reuse existing lyrics. Describe a fictional performer and the style of music, and only animate faces you have the right to use.

How many credits does it cost?

It depends on the model, resolution and length. On this page, a 5-second Wan 3.0 clip at 480p was 48 credits, the 10-second Seedance 2.5 clip at 720p was 740, the 15-second shot list was 494, and the song-first singer took 214 (song 12, photo 51, lip sync 151). The price is shown on the button before you generate.