Using MiniMax Speech 2.8 HD in your workflow

Sorceress brings MiniMax models into a shared API catalog. Select this model explicitly, check its supported inputs, and keep your API key on your server. Model capabilities differ; the schema and options on this page describe this entry rather than every model from the same brand.

Add sound or speech to a creative workflow.
Integrate audio generation with your application using the model-specific parameters.

Pricing and availability

Cost: 1 credit per 2,000 characters (min 1). Charged only on success.. Successful work is billed in Sorceress credits. Defaults and optional inputs can affect the total.

Inputs and supported options

textstring · required

Text to speak (max 10,000 chars).

voice_idstring · required

A preset voice id (e.g. "Wise_Woman") or one of the caller's cloned voice ids. Use speech_list_voices to discover both.

speednumber · optional

Speech rate 0.5–2.0 (default 1).

pitchnumber · optional

Pitch -12 to 12 (default 0).

emotionstring · optional

One of: none, happy, calm, sad, angry, fearful, disgusted, surprised (default none).

volumenumber · optional

Volume multiplier (default 1).

language_booststring · optional

Optional language hint to improve pronunciation (e.g. "English", "Spanish").

Workflow actions

Speech (TTS)

Convert text to spoken audio with a preset or cloned voice. Async: returns a jobId — poll GET /api/v1/jobs/:id until it succeeds, then read result.audio. Call speech_list_voices first to get valid voice_id values.

Cost: 1 credit per 2,000 characters (min 1). Charged only on success.

List Speech Voices

List the voices available to speech_generate for this account: 17 preset voices plus the caller's own cloned voice ids. Returns immediately (not a job).

0 credits per successful call

API request template

Replace placeholders and choose valid values from the input contract.

POST /api/v1/tools/speech_generate
Authorization: Bearer $SORCERESS_API_KEY
Content-Type: application/json

{
  "text": "<text>",
  "voice_id": "<voice_id>"
}

Read the endpoint reference →

Completion, results and billing

Save the job identifier returned by asynchronous requests and poll the jobs endpoint until the job succeeds or fails. Do not resubmit a generation just because it is still processing. Chat returns a completed reply and usage details. A multi-step workflow can include separately billed successful steps even if a later step fails. Keep returned asset links with your project, and consult each action for its output format.

Coding agents can discover the current catalog and input requirements from the Sorceress agent instructions. Use a Sorceress key; provider credentials stay on the server.