skills.rest
genmedia-labs
genmedia-labsOFFICIAL
Explore all skills by this author →
GitHub

ace-step

Generate, inpaint, and outpaint music with ACE Step via the RunComfy CLI.

12|2|Updated Aug 12, 2026
Search Tags:#music generation#text-to-audio#ai music#runcomfy#ace step#audio inpainting#audio outpainting
One-click install
npx skills add https://github.com/genmedia-labs/skills --skill ace-step-genmedia-labs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ace-step
Source: https://github.com/genmedia-labs/skills/tree/main/ace-step
Command: npx skills add https://github.com/genmedia-labs/skills --skill ace-step-genmedia-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @runcomfy/cli.

What problem does it solve? Creating original music, fixing a bad section of a track, or extending a short clip into a full song normally requires expensive commercial music APIs or manual audio editing. This Skill gives you tag-driven music generation, time-range inpainting, and bidirectional outpainting through four ACE Step endpoints on RunComfy at $0.0002–0.0003 per second of audio. ## Core Features & Use Cases - Text-to-audio generation: Compose 5–240 second stereo tracks from comma-separated genre, mood, and instrument tags, with optional structured lyrics using [Verse]/[Chorus]/[Bridge] markers. The ACE Step 1.5 variant supports vocals in 50+ languages. - Audio inpainting: Regenerate a specific time range inside an existing track (e.g., replace a bad chorus from 20–40 s) using start_time/end_time anchors relative to track start or end. - Audio outpainting: Extend a track bidirectionally by adding an intro before and/or an outro after an existing clip, up to 4 minutes total. - Use Case: Draft a 30-second hook with the cheap base model, then use audio-outpaint to add a 30-second intro and 60-second outro, producing a full 2-minute cut in one workflow. ## Quick Start Ask the agent to generate a 60-second lo-fi hip-hop instrumental track with ACE Step and save it to the ./out folder.

Frequently Asked Questions about ace-step

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate music with ACE Step from text tags?▼

Run the runcomfy CLI with the acestep-ai/ace-step/text-to-audio endpoint and pass comma-separated genre, mood, and instrument tags in the JSON input. Set duration between 5 and 240 seconds and use "[inst]" in the lyrics field for instrumental output.

ACE Step vs ElevenLabs Music: which should I use?▼

ACE Step costs $0.0002–0.0003 per second, roughly 27x cheaper than ElevenLabs Music, and supports inpainting and outpainting. ElevenLabs Music suits polished commercial vocal cuts, while ACE Step fits cost-sensitive drafts, batches, and open-weights pipelines.

How do I fix a bad section of an existing song with AI?▼

Use the acestep-ai/ace-step/audio-inpaint endpoint with the source audio URL and start_time/end_time in seconds to define the editable range. The model regenerates only that segment using your tags, and times can be anchored relative to the track start or end.

Does ACE Step support multilingual lyrics?▼

Yes, the ACE Step 1.5 text-to-audio endpoint supports vocals in over 50 languages. Write lyrics directly in the target language, include the language in your tags, and use section markers like [Verse] and [Chorus] for structure.

What is the maximum audio length ACE Step can generate?▼

Each text-to-audio call produces 5 to 240 seconds (4 minutes) of stereo audio. For longer tracks, chain multiple calls or use audio-outpaint to extend an existing clip, keeping the combined total under 4 minutes per call.

Why is my ACE Step generation not matching the style I want?▼

Tags drive composition, so vague tags like "chill music" produce generic results. Use specific comma-separated tags including genre, mood, instruments, and BPM, and fix the seed parameter to reproduce results you like.

Related Skills

Explore More Design & Creative Skills→
duotifyduotify
252

gemini-audio-transcriber

Transcribe audio files into Traditional Chinese Markdown transcripts with speaker labels.

Official
Advanced
DaytimeBluesDaytimeBlues

game-audio

Guide game audio design covering SFX, music, adaptive systems, and 3D spatialization.

Community
Intermediate
GeneralReasoningGeneralReasoning
4

FFmpeg Audio Processing

Extract, normalize, and mix audio tracks from video files using FFmpeg.

Official
Intermediate
runcomfy-comruncomfy-com
12

ai-music

Generate and edit AI music via RunComfy with selected audio models.

Official
Advanced