ACE Step 1.5 Music Generator

Generate full songs up to 4 minutes from style tags and optional lyrics. Supports 50+ languages, structured song sections, and instrumental-only mode.

Cost: 3 credits

Input

Comma-separated genre, instrument, mood, and tempo tags to define the musical style.

Optional. Use section markers like [Verse], [Chorus], [Bridge]. Leave empty for instrumental tracks.

Track length in seconds. Maximum 4 minutes (240s).

Use a specific seed for reproducible results. -1 generates a random seed each time.

Output

Generated content will appear here

Frequently Asked Questions

What are style tags and how should I write them?
Style tags are comma-separated keywords that describe the genre, instruments, mood, and tempo of your song. For example: 'pop, female vocals, upbeat, guitar, 120bpm' or 'lo-fi, chill, ambient, piano, rainy mood'.
How do I format the lyrics?
Use section markers in square brackets to structure your song: [Verse], [Chorus], [Bridge], [Intro], [Outro]. Each marker starts a new section. Leave the lyrics field empty to generate instrumental-only tracks.
What is the maximum track duration?
You can generate tracks up to 240 seconds (4 minutes). The default duration is 60 seconds.
Can I generate instrumental music without vocals?
Yes! Simply leave the lyrics field empty, or include [inst] or [instrumental] in the lyrics field. The model will generate pure instrumental music based on your style tags.
What languages are supported for lyrics?
ACE Step 1.5 supports over 50 languages for lyrics, including English, Chinese, Japanese, Spanish, French, and many more.
How can I reproduce the same result?
Use a specific seed value (any positive integer). The same tags, lyrics, duration, and seed will produce an identical output. Set seed to -1 for random generation each time.