Developers

Seed-Audio 1.0 API access, mapped.

There is no public SDK and no open weights. There are three legitimate ways to call the model today, and they differ in onboarding friction more than in capability.

The three routes

BytePlusFirst-party international. $0.15 per output minute, 300 free minutes on activation. Business verification required. Best unit economics at volume

Volcano EngineFirst-party China. Invited beta via the Ark console; requires a China-verified account. Production pricing unpublished

fal.aiLicensed partner, self-serve signup, $0.1875 per output minute. Fastest way to a working key for individuals

The request shape

All first-party routes share the same core contract: a synchronous POST that returns base64 audio. The BytePlus form looks like this:

POST /api/v3/tts/create
X-Api-Key: <your key>
Content-Type: application/json

{
  "model": "seed-audio-1.0",
  "text_prompt": "A 1940s radio drama cold open...",
  "audio_config": { "format": "mp3", "sample_rate": 48000 }
}

// response: { "audio": "<base64>", "original_duration": 15.1 }

Billing follows original_duration, by the second.
Takes are capped at 120 seconds; chain takes for long-form.
Reference audio for cloning: up to 3 clips of 30s, cited as @Audio1 to @Audio3 in the prompt.
Explicit and metadata watermark flags exist and default to off on the first-party API.

Before you build on a relay

Aggregators resell access below official rates, sometimes by billing input characters instead of output minutes. That pricing inverts their margin on exactly the prompts this model is best at, so treat any too-cheap rate as temporary. If uptime matters, hold a first-party key as fallback. The full math is on the pricing page.

A unified audio API

We are building one endpoint that fronts Seed-Audio and other scene and speech models with a single request shape and per-second billing. It ships after credit packs. Until then, the playground is the fastest way to evaluate output quality without any onboarding.