Seed Audio
Seed Audio is ByteDance's voice and audio-scene model. Use it for dialogue, narration, character voice scenes, ambience, and sound design from text, audio references, or an image reference.
Best for
- voice scenes
- dialogue
- audio references
- sound design
Use when
- You want generated voice or dialogue with scene-level audio direction.
- You have audio references that should guide voice, tone, or delivery.
- You need ambience or sound design around the spoken audio.
- You need Seed Audio-specific sample rate or output format controls.
Avoid when
- You need a reusable voice-library or cloning workflow; use ElevenLabs.
- You need a normal generated music track; use Lyria or ElevenLabs Music.
Inputs
- audio prompt
- optional audio references
- optional image inspiration
Outputs
- audio
DayGen surfaces
- /music/create