● Connecting

MADE FOR YOUR MOMENT

Your sound.
Press play.

Your collection, your discoveries, your next creation.
All your sounds in one place.

ON REPEAT

Most played

Listening across the community

FIND YOUR FREQUENCY

Explore genres

YOUR COLLECTION

My library 0

TRACK / ARTISTGENREDURATION

SOLMUSIC STUDIO

One idea.
Endless possibilities.

Describe what you want to hear. ACE-Step brings it to life.

Track title · optionalWhat should your music sound like?Lyrics · optional for instrumental tracks

Duration in secondsModelLanguage

SHAPE YOUR SOUND

Reference audio

Optional

Upload a track to guide the sound and style. This is a reference for new music, not a cover or an exact copy.

♫ Choose reference audio

WAV, MP3, FLAC, OGG, OPUS, M4A, or AAC · up to 25 MB. The file is sent to ACE-Step with your request.

⚙ Advanced settingsTempo, key, seed, sampling, and AI planning⌄

Defaults are ready to use. These controls are sent to the installed ACE-Step API.

Musical direction

Tempo (BPM)Leave blank to let ACE-Step choose. Range: 30–300.Musical keySuggest a tonal center and major/minor mood.Time signatureBeats per bar. Musical guidance, not a guarantee.

Generation

Inference stepsStandard: 1–200, default 50. More steps take longer.Prompt guidanceHigher values push the model toward your description. Excessive guidance can reduce quality.SamplerChanges the diffusion sampling method.

Variations1–4 outputs per request. More variations use more GPU memory.Output formatFormat saved by the generator. Registered tracks can be downloaded as MP3 by their creator.SeedBlank = a fresh seed. A fixed number helps compare settings; results may still vary with the environment.

Synced lyrics

Generate synced lyricsACE-Step aligns the supplied lyrics after generating audio. Adds processing time and GPU memory. Instrumental tracks are skipped. If alignment fails, your audio is preserved.

AI planning

Optional: these settings require ACE-Step’s auxiliary 5Hz language model. Leave them off when the service runs with the language model disabled.

Plan audio with the language modelGenerates audio codes before diffusion. Adds processing time.Allow description refinementLets the language model develop your music description.Allow vocal language detectionLets the language model infer language from your lyrics. Turn off to keep the selected language.

Planning creativityLanguage model temperature. Higher = more variation.Planning top-pLimits the pool of likely language-model tokens. Default: 0.9.

ACE-Step must be running before you generate. SolMusic does not automatically start the service.

♫

Your next sound is hereChoose a track to get started

0:000:00

Your sound.
Press play.

Most played

Explore genres

My library 0

One idea.
Endless possibilities.

Reference audio

Musical direction

Generation

Synced lyrics

AI planning

Play queue

Lyrics

Create with SMUSIC

Additional tracks

Your musician / band profile

Pending payments

Manage track

Author credit

Visibility

Pay with SMUSIC

Delete track?

Organize track

Sign in

Enable two-factor authentication

Disable two-factor authentication

Your sound.Press play.

Most played

Explore genres

My library 0

One idea.Endless possibilities.

Reference audio

Musical direction

Generation

Synced lyrics

AI planning

Play queue

Lyrics

Your sound.
Press play.

One idea.
Endless possibilities.