Skip to main content

ByteDance

Seed Audio 1.0 AI text to speech

Speech in a voice you describe or attach

Seed Audio 1.0 is ByteDance’s speech model, available on Ochra for text-to-speech. It reads your script in a voice you describe or in a voice from a recording you attach with the speaker’s permission, and ByteDance describes it keeping a voice recognisable across languages.

Released by ByteDance: July 20, 2026On Ochra since: September 1, 2026

Seed Audio 1.0 price per script

Prices as of September 28, 2026, in credits, with an approximate dollar figure at the per-credit price of the smallest credit pack. The composer shows the same price before you generate.

Seed Audio 1.0 — priced by the length of the script
ScriptPrice
200 characters3 credits≈ $0.17
1,000 characters12 credits≈ $0.69
3,000 characters12 credits≈ $0.69

What it takes

Voice recordings
up to 3

What is different about Seed Audio 1.0

  • Takes reference recordings of a voice, which the MiniMax speech rows here do not.
  • ByteDance describes a voice kept recognisable when the script changes language.
Strengths and limits

Key Features

  • Speak in a voice from a recording you are entitled to use, without training a model for it.
  • Priced by the length of the script, so a short line costs little.

Where it falls down

  • A recording must be of your own voice or used with the speaker’s permission, and you confirm that when you attach it.
  • The number of reference recordings per request is capped; the limit is listed on this page.

Seed Audio 1.0 capabilities

Seed Audio 1.0 — Model family
ModelToolsDurationResolutionAudio
Seed Audio 1.0Text-To-Speech——

Seed Audio 1.0 questions

Yes. Attach a recording of your own voice, or one you have the speaker’s permission to use, and the model reads the script in that voice. The confirmation is recorded with the generation.

Other Seed Audio releases

Ready to Get Started?

Every tool shows its price in credits before you run it.