Stable Audio 3 Medium (Text-to-Audio)

Stable Audio 3 Medium makes stereo music up to 6 minutes 20 seconds long from a text prompt, as MP3, WAV or FLAC, for a flat $0.0376 a clip.

Stable Audio 3 Medium is Stability AI's 1.4-billion-parameter text-to-audio model, released in May 2026. It makes stereo music up to 6 minutes 20 seconds long from a text prompt and was trained on licensed data. On the Artificial Analysis instrumental music arena it stood 14th of 19 models in November 2026; it has no entry on the vocal board.

Use it for background music, loops, intros and beds: describe the genre, mood, instruments and tempo in prompt, and what to keep out in negative_prompt. duration sets the length, 1 to 380 seconds (30 by default); seed repeats a result; num_inference_steps (1-100, 8 by default) rarely needs changing on this distilled model; output_format picks mp3 (the default, 192 kbit/s), wav, flac, ogg, opus, m4a or aac.

Inputs: prompt (required), negative_prompt, duration, seed, num_inference_steps and output_format. Outputs: audio (the track) and seed.

Price: $0.0376 per audio at list, a flat price: the length does not change it. It makes one audio per request. It takes no lyrics (use MiniMax Music 3 or Mureka V9.5 for songs with vocals) and no reference track (the Audio Inpainting and Outpainting models rework a track you have), and it makes no speech.

Price: $0.0376 per run. One run with default settings costs $0.0376.

Run Stable Audio 3 Medium (Text-to-Audio) via API

curl -X POST https://api.flymy.ai/api/v1/flymyai/stable-audio-3-medium-text-to-audio/predict \
  -H "X-API-KEY: $FLYMYAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"prompt":"A calm lo-fi piano loop with soft vinyl crackle","duration":5}'

API documentation · Create a free account · Pricing

Related models