Model API

Stable Audio 3 Small Sfx Base

Stable Audio 3 Small SFX Base audio inpainting is the foundational 459 million parameter checkpoint for editing or filling selected sound-effect segments guided by text prompts.

Model id

Use this in every request

stable-audio-3/small/sfx/base/audio-inpainting

This model has other modes with their own ids: stable-audio-3/small/sfx/base/text-to-audio, stable-audio-3/small/sfx/base/audio-to-audio, stable-audio-3/small/sfx/base/audio-outpainting.

Run it

Quote the input, start the generation with the quote, then poll the job. Set RAYWAKE_API_KEY first; see Authentication.

# 1. Price the exact input (quotes are valid for up to 10 minutes)
curl -s https://api.raywake.com/v1/quotes \
  -H "Authorization: Bearer $RAYWAKE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "stable-audio-3/small/sfx/base/audio-inpainting",
  "input": {
    "prompt": "A red ceramic cup on a wooden table, soft morning light",
    "audio_url": "https://example.com/input.mp3",
    "mask_start_seconds": 0,
    "mask_end_seconds": 0
  }
}'

# 2. Start the generation with the quote_id from step 1
curl -s https://api.raywake.com/v1/generate \
  -H "Authorization: Bearer $RAYWAKE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{"model": "stable-audio-3/small/sfx/base/audio-inpainting", "quote_id": "QUOTE_ID", "input": {"prompt":"A red ceramic cup on a wooden table, soft morning light","audio_url":"https://example.com/input.mp3","mask_start_seconds":0,"mask_end_seconds":0}}'

# 3. Poll until the status is terminal
curl -s https://api.raywake.com/v1/jobs/JOB_ID \
  -H "Authorization: Bearer $RAYWAKE_API_KEY"

Input

Input fields for stable-audio-3/small/sfx/base/audio-inpainting
Name	Type	Description
`negative_prompt`	string	Text description of qualities to avoid in the output.
`num_inference_steps`	integer	Number of sampling steps. Base (non-distilled) checkpoints typically need ~50 for good quality. Default: `50`
`guidance_scale`	number	Classifier-free guidance scale. Higher values follow the prompt more strictly; ~7.0 is a good starting point for base checkpoints. Default: `7`
`seed`	integer	Random seed for reproducible outputs. Omit for a random seed.
`enable_prompt_expansion`	boolean	If True, the prompt will be expanded using an LLM for more detailed and higher quality results. Default: `false`
`sync_mode`	boolean	If True, the audio is returned inline as a data URI and the result is not saved to the request history. Default: `false`
`output_format`	string (enum)	Container format for the generated audio output. Default: `mp3` One of: `mp3`, `wav`, `flac`, `ogg`, `opus`, `m4a`, `aac`
`bitrate`	string	Audio bitrate for compressed output formats (e.g., mp3, aac, opus). Format e.g. '192k' or '320k'. Ignored for lossless formats (wav, flac). Default: `192k`
`prompt`required	string	Text description guiding how the masked region is regenerated.
`audio_url`required	string (URL)	Source audio. Only the masked region is regenerated.
`mask_start_seconds`required	number	Start of the region to regenerate, in seconds.
`mask_end_seconds`required	number	End of the region to regenerate, in seconds.

Output

A succeeded job lists its results under outputs, stored on our side. See Jobs for every field.

Job excerpt

"outputs": [
  { "type": "audio", "url": "https://…", "mime_type": "audio/mpeg", "thumbnail_url": null }
]

Pricing

up to, default options$0.06/ run

The exact charge for your input comes from a quote. How charges settle