Model Information

Display Name: Stable Audio 3.0

API Model ID: stability-ai/stable-audio-3

Category: Text To Audio

Description: Stable Audio 3.0 is Stability AI's most advanced audio model — up to 6-minute, 44.1 kHz stereo compositions from text or existing audio. This is an ASYNC service: it returns a generation id to be polled at audio/results/{id}. **Key Features:** - Up to 6-minute, 44.1 kHz stereo output - Fast generation (4-8 steps) - Text-to-audio and audio-to-audio (input audio via images[]) - Pre-trained on licensed data (AudioSparx + Freesound) **Best For:** - Long-form music production and cinematic sound design - Remixing and extended compositions **Note:** Asynchronous — poll audio/results/{id} for the final audio.

How to Use This Model

To use Stable Audio 3.0 via the HInow.ai API, use the model ID: stability-ai/stable-audio-3

API Request Example (Chat/Text)

POST https://api.hinow.ai/v1/chat/completions
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
  "model": "stability-ai/stable-audio-3",
  "messages": [
    {"role": "user", "content": "Your message here"}
  ]
}
              

Pricing

• total: $0.39

Available Parameters

• duration: Duration of the generated audio in seconds (1-380).
• steps: Number of sampling steps (4-8).
• guidance_scale: How strictly the result adheres to the prompt (cfg_scale). Maps to the provider's cfg_scale.
• strength: Audio-to-audio denoising strength (0-1): 0 keeps the input audio, 1 ignores it. Only used when an input audio is provided via images[].
• output_format: Output audio format. (Options: mp3, wav)
• seed: Random seed for reproducibility. -1 (or omit) for a random seed.

Quick Reference

To use this model, set: "model": "stability-ai/stable-audio-3"

Featured: No

Documentation: https://hinow.ai/models/stability-ai/stable-audio-3

API Endpoint: https://api.hinow.ai/v1