Music Video

Turn a song into a finished music video with an AI Director agent — you make every key creative decision.

What Is Music Video

Music Video turns a song into a finished, fully edited music video. You work with an AI agent — the Director — that handles scriptwriting, casting, scene design, choreography, storyboarding, video generation, and editing. You stay in charge of the creative decisions.

Your Director

The Director is a single agent that runs the whole production end to end, and you're the only person it answers to. Each phase of production is driven by a dedicated skill — a professional playbook the Director picks up when that phase begins, the way a film director switches between department briefs. You'll sometimes see the Director reading or using a skill in the chat; that's all there is to it — skills work on their own, and you never need to manage them.

What that looks like across a production:

Phase	What the Director does	What you see
Creative planning	Writes the creative plan and the script for every shot	Creative plan, shot scripts
Casting	Designs your character's base look and per-scene looks	Character sheets (multi-view), per-scene looks
Art & storyboards	Defines the visual style, designs scene environments, draws storyboards	Style guide, scene references, storyboards
Choreography	Designs dance and performance moves, creates motion keyframe grids	Choreography docs, motion keyframes
Cinematography	Writes generation prompts, runs video generation, handles lip-sync refinement	The final video shots

How It Works

1. Start from your ACE Studio project — a song you produced, or a finished track you imported. The Director renders it as your song, analyzes it, and asks you four questions: visual style, storytelling approach, lead character, and aspect ratio.

2. The Director works through the pipeline — creative plan, casting, scene design, shot scripts, looks, optional choreography, and storyboards. You review and give feedback at each milestone.

3. The storyboards are placed on the Canvas as an animatic, so you can preview the full video's pacing before anything is rendered.

4. When everything is ready, you give the go-ahead. Shots render in parallel, replace the storyboards automatically, and singing shots get a lip-sync refinement pass.

5. Watch the result, then request changes — only the affected shots are regenerated.

For the full walkthrough, see Creating a Music Video. For revisions, manual edits, and resuming a project, see Editing & Iteration.

Credit Pricing

Music Video uses Credits for music analysis, image generation, video generation, and Agent LLM usage.

Category	Model or feature	Billing unit	Credit cost
Music analysis	Music Analysis	Per request	≈20 per minute
Image models	Midjourney	Per request	20
Image models	Nanobanana	Per request	25
Image models	GPT Image 2	Per request	15
Image models	Qwen Image Multiple Angles	Per request	7
Video models	Seedance 2 (1080p)	Per second	68
Video models	Seedance 2 (720p)	Per second	34
Video models	Lipsync 3	Per second	25
Agent	Agent LLM Tokens	Based on actual token usage	Varies by task