Seed Audio 1.0

Seed Audio 1.0 is ByteDance's all-in-one audio generation model. From one text prompt it produces voice, instrumental music, and sound effects together as a finished, mixed track. It also edits existing audio: extend a clip, fill a gap, swap a line, or stitch two takes.

Seed Audio 1.0 clones a voice zero-shot from up to three reference clips of about 30 seconds each, with no training or fine-tuning. The cloned voice keeps its accent, tone, and character across the whole generation. You can also define a voice from a text description or a character image instead of a recording.

Yes. Write a scene with several characters and label each line, for example Host: ... and Guest: ... . Seed Audio 1.0 gives each speaker a distinct voice, emotion, and pacing in a single generation.

Seed Audio 1.0 generates up to two minutes of audio in a single pass. Continuation mode extends it further while keeping the voice character and style consistent with what came before.

Seed Audio 1.0 supports English and Chinese, with broader language support planned. For voice cloning, matching the reference clip language to the output language gives the most consistent result.

Text-to-speech turns text into a single voice track. Seed Audio 1.0 generates the whole scene, the voice, background music, and sound effects together in one output, and can revise specific sections afterward. The difference is scope: a finished audio production versus only the voice.