Music Video Montage: Turn audio into a music video in minutes

Got a track? Use Music Video Montage to turn it into a full music video. Drop in your audio, set a few options, and Music Video Montage does the rest. It analyzes your audio creates a video based on the style, content and emotions in your music. No editing required.

Three Modes to generate with

Generate: Create a music video from your audio and description
Reference: Create a music video from your audio, description with a reference image
Restyle: Create a music video from your audio, description a inspirational guiding image

Access Music Video Montage

Click Music Video Montage in the Dashboard workflows, or open Canvas and open a Create Video Flow. Click the model name to open the model menu and scroll down to choose Music Video Montage.

Add your Audio

Upload your audio file from your Media Library or device. Video Montage analyzes it directly, so the video it generates is shaped by the content and mood of the sound.

Prompting for Music Video Montage

It's a montage, not a story. One character, one world, one style, caught mid-moment. The system handles camera, pacing and motion on its own — your job is just who, where, and how it looks.

Three things to include in your prompt

Subject — who or what, doing something (not standing still). Give fixed physical detail: hair, build, clothing, texture. Skip race/nationality labels, describe the features instead.

Location — one grounded setting: time of day, weather, a texture or two. Let it carry the mood rather than naming it.

Style — one clear keyword up front: anime, oil painting, 1970s film photograph, claymation. Vague or hedged style descriptions don't lock as well.

Keep it specific, not long

"A woman singing" tells the model nothing. "A woman with close-cropped silver hair in a weathered leather jacket, standing in a rain-slicked alley" gives it something to work with.

Leave out

• Story language — "first," "then," "as the song builds"

• Camera direction — zoom, pan, dolly, handheld (already automated)

• Weapons, injury or drug specifics — use atmosphere instead

• More than one competing subject

No style opinion? No problem

A short subject line with the style field left blank works fine — the audio analysis will pick a genre-appropriate look. This works best on songs with a strong, clear genre identity.

Prompt Template

Subject: [description] + [mid-action] + [texture detail] Style: [one keyword]

Example: A weathered fisherman with sun-cracked skin, hauling a net onto a wooden dock at dusk, salt-spray catching the last light / Style: 1970s Kodachrome film photograph

Test wording at low quality before running a high-quality pass.

Settings

Add reference image (for Restyle and Reference modes)

Change from “Generate” mode to “Reference” or “Restyle” to add an image.

Reference uses the image you upload as a guide for each clip. Restyle uses your aesthetic prompt and sound to restyle the image to create your clips.

Choose your video length, aspect ratio and resolution

Video length: Choose 1, 2, or 3 minutes.

Aspect ratio: 16:9, 9:16, 1:1, 3:4, or 4:3

Resolution: 540, 720, or 180p

Multi Clip toggle on or off

Want more visual variety within each clip? Turn on Multi Shot. Instead of one continuous shot per clip, you’ll get multiple shots packed into each one. This makes a big difference with longer videos, as the number of clips in each video stays the same, with longer clips for longer videos.

Generate your music video

Once you’re happy with your prompt and settings, click the Generate button to create your video.

Find out what's happening behind the scenes when you create your video, how your audio is analyzed, and more on how to prompt for best results:
Music Video Montage: Full Prompt Guide