ace/ace-step/audio-to-audio
Ace Step Audio To Audio by ace - advanced AI model for audio-to-audio. Delivers high-quality results with fast inference, suitable for both creative and production workflows.
Input
URL of the audio file to be outpainted.
Lyrics to be sung in the audio. If not provided or if [inst] or [instrumental] is the content of this field, no lyrics will be sung. Use control structures like [verse], [chorus] and [bridge] to control the structure of the song.
0200
Guidance scale for the generation. Range: 0 to 200.
Random seed for reproducibility. If not provided, a random seed will be used.
0200
Minimum guidance scale for the generation after the decay. Range: 0 to 200.
010
Tag guidance scale for the generation. Range: 0 to 10.
010
Lyric guidance scale for the generation. Range: 0 to 10.
Whether to edit the lyrics only or remix the audio. Allowed values: lyrics, remix.
01
Guidance interval for the generation. 0.5 means only apply guidance in the middle steps (0.25 * infer_steps to 0.75 * infer_steps) Range: 0 to 1.
Original seed of the audio file.
Scheduler to use for the generation process. Allowed values: euler, heun.
-100100
Granularity scale for the generation process. Higher values can reduce artifacts. Range: -100 to 100.
Type of CFG to use for the generation process. Allowed values: cfg, apg, cfg_star.
Original lyrics of the audio file.
01
Guidance interval decay for the generation. Guidance scale will decay from guidance_scale to min_guidance_scale in the interval. 0.0 means no decay. Range: 0 to 1.
360
Number of steps to generate the audio. Range: 3 to 60.
Idle
Example output — click Run to generate your own
Production endpoint
Send your first request
OpenAI-compatible endpoint with unified authentication and usage tracking.
POST
https://api.sandbase.ai/v1/runModel ID
ace/ace-step/audio-to-audio01
Input Schema
18 parameters · 0 required · 18 optional
| Parameter | Type | Required | Description |
|---|---|---|---|
seed | integer | Optional | Random seed for reproducibility. If not provided, a random seed will be used. |
tags | string | Optional | Comma-separated list of genre tags to control the style of the generated audio. |
audio | string | Optional | URL of the audio file to be outpainted. |
lyrics | string | Optional | Lyrics to be sung in the audio. If not provided or if [inst] or [instrumental] is the content of this field, no lyrics will be sung. Use control structures like [verse], [chorus] and [bridge] to control the structure of the song. · Default: "" |
edit_mode | string | Optional | Whether to edit the lyrics only or remix the audio. · Options: lyrics, remix · Default: "remix" lyricsremix |
scheduler | string | Optional | Scheduler to use for the generation process. · Options: euler, heun · Default: "euler" eulerheun |
guidance_type | string | Optional | Type of CFG to use for the generation process. · Options: cfg, apg, cfg_star · Default: "apg" cfgapgcfg_star |
original_seed | integer | Optional | Original seed of the audio file. |
original_tags | string | Optional | Original tags of the audio file. |
guidance_scale | number | Optional | Guidance scale for the generation. · Min: 0 · Max: 200 · Default: 15 |
number_of_steps | integer | Optional | Number of steps to generate the audio. · Min: 3 · Max: 60 · Default: 27 |
original_lyrics | string | Optional | Original lyrics of the audio file. · Default: "" |
granularity_scale | integer | Optional | Granularity scale for the generation process. Higher values can reduce artifacts. · Min: -100 · Max: 100 · Default: 10 |
guidance_interval | number | Optional | Guidance interval for the generation. 0.5 means only apply guidance in the middle steps (0.25 * infer_steps to 0.75 * infer_steps) · Min: 0 · Max: 1 · Default: 0.5 |
tag_guidance_scale | number | Optional | Tag guidance scale for the generation. · Min: 0 · Max: 10 · Default: 5 |
lyric_guidance_scale | number | Optional | Lyric guidance scale for the generation. · Min: 0 · Max: 10 · Default: 1.5 |
minimum_guidance_scale | number | Optional | Minimum guidance scale for the generation after the decay. · Min: 0 · Max: 200 · Default: 3 |
guidance_interval_decay | number | Optional | Guidance interval decay for the generation. Guidance scale will decay from guidance_scale to min_guidance_scale in the interval. 0.0 means no decay. · Min: 0 · Max: 1 · Default: 0 |
02
Output Schema
| Field | Type | Description |
|---|---|---|
id | string | Unique identifier for the generation task |
status | string | Task status: pending, running, completed, failed, timeout |
model | string | Model used for the generation |
outputs | array | Array of output items |
outputs[].url | string | URL of the generated artifact |
outputs[].content_type | string | MIME type (e.g. image/png, video/mp4) |
error | object | null | Error details if failed, null on success |
error.type | string | Machine-readable error type code |
error.message | string | Human-readable error description |
Async Workflow
This model uses asynchronous execution. Submit a request and poll for the result.
- Submit — POST to /v1/run, receive an
id - Poll — GET /v1/run/{id} until status is
completed,failed, ortimeout - Retrieve — Read
outputsfrom the completed response
03
Code Examples
Ready-to-run snippets
# Step 1: Submit
curl -X POST https://api.sandbase.ai/v1/run \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "ace/ace-step/audio-to-audio",
"tags": "lofi, hiphop, drum and bass, trap, chill",
"audio": "https://static.sandbase.ai/examples/ace/ace-step/audio-to-audio/input_audio_0.wav",
"lyrics": "",
"edit_mode": "remix",
"scheduler": "euler",
"guidance_type": "apg",
"original_tags": "lofi, hiphop, drum and bass, trap, chill",
"guidance_scale": 15,
"number_of_steps": 27,
"original_lyrics": "",
"granularity_scale": 10,
"guidance_interval": 0.5,
"tag_guidance_scale": 5,
"lyric_guidance_scale": 1.5,
"minimum_guidance_scale": 3,
"guidance_interval_decay": 0,
"prompt": "a beautiful sunset over mountains"
}'
# Step 2: Poll result (replace <id>)
curl https://api.sandbase.ai/v1/run/<id> \
-H "Authorization: Bearer YOUR_API_KEY"API README
ACE-Step
ACE-Step Audio-to-Audio transforms an existing recording through either remix or lyrics-focused editing. The source audio supplies musical structure and identity, while new tags, lyrics, and generation controls guide the revised result.
Remix mode is suited to broader stylistic reinterpretation, while lyrics mode focuses the request on changing vocal content. Original tags, lyrics, and seed can be supplied alongside the new direction, giving the model more context for preserving or intentionally departing from the source.
Highlights
Two editing modes. Choose a broad remix or a lyrics-focused transformation according to the creative goal.
Source-conditioned generation. Builds on an existing recording instead of composing without musical context.
Original-versus-new guidance. Accepts original tags, lyrics, and seed together with replacement direction.
Fine generation control. Scheduler, steps, sampling seed, and separate tag and lyric guidance support deliberate revisions.
Pricing
| Billing unit | Price |
|---|---|
| Per generated second | $0.0002 |
When to Use
| ✅ Good fit | ❌ Consider alternatives |
|---|---|
| The model's named workflow matches the source material and intended output | A different input modality or model route is required |
| A managed asynchronous result is suitable for the production pipeline | A synchronous, interactive editor is essential |
| The documented controls cover the required duration, framing, or format | The project needs controls outside this endpoint's schema |
| Creative iteration benefits from a repeatable request structure | Exact deterministic pixels, frames, geometry, or samples are mandatory |
| A finished downloadable media asset is the desired deliverable | Editable source layers or a native project file are required |
Prompt Guide
For music generation, state the intended result first, then add the subject or source treatment, progression, style, and delivery constraints. Keep one creative variable per phrase, use the documented field names for controls, and change one setting at a time when comparing results.
{
"audio": "https://static.sandbase.ai/examples/ace/ace-step/audio-to-audio/input_audio_0.wav"
}
Technical Specs
| Spec | Value |
|---|---|
| Model ID | ace/ace-step/audio-to-audio |
| Inputs | audio, edit_mode, granularity_scale, guidance_interval, guidance_interval_decay, guidance_scale, guidance_type, lyric_guidance_scale, lyrics, minimum_guidance_scale, number_of_steps, original_lyrics, original_seed, original_tags, scheduler, seed, tag_guidance_scale, tags |
| Required inputs | None |
| Output fields | content_type, url |
| Execution | Async (submit, then poll for result) |
Related Models
ace/ace-step— Compare a nearby route in the same local model family.ace/ace-step/audio-inpaint— Compare a nearby route in the same local model family.ace/ace-step/audio-outpaint— Compare a nearby route in the same local model family.