Text to Speech

CheapAI exposes multiple text-to-speech flows behind one account. Choose the route that matches your voice source, model family, and output controls.

Choose The Right Route

Voice-based TTS

POST /v1/text-to-speech/:voice_id

Use when you want a specific voice id and standard TTS output formats.

MiniMax TTS

POST /v1m/task/text-to-speech

Use when you want MiniMax model selection, voice_setting controls, or MiniMax voice inventory.

Edge TTS

POST /v1e/task/text-to-speech

Use when you need broad voice coverage, SRT input, or word-level transcript output.

Voice-Based TTS

http

POST https://api.cheapaiapi.com/v1/text-to-speech/{voice_id}?output_format=mp3_44100_128

textstringrequiredInput text to synthesize.

model_idstringTTS model, for example eleven_multilingual_v2.

with_transcriptbooleanInclude transcript outputs when supported.

receive_urlstringOptional webhook endpoint for completion delivery.

bash

curl -X POST "https://api.cheapaiapi.com/v1/text-to-speech/21m00Tcm4TlvDq8ikWAM?output_format=mp3_44100_128" \
  -H "Authorization: Bearer sk_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "The first move is what sets everything in motion.",
    "model_id": "eleven_multilingual_v2",
    "with_transcript": false
  }'

MiniMax TTS

http

POST https://api.cheapaiapi.com/v1m/task/text-to-speech

textstringrequiredText to synthesize.

modelstringMiniMax TTS model such as speech-2.6-hd or speech-2.6-turbo.

voice_setting.voice_idstringrequiredVoice id from the MiniMax voice list.

voice_setting.volnumberVolume control.

voice_setting.pitchnumberPitch control.

voice_setting.speednumberSpeed control.

language_booststringLanguage hint from the common config endpoint.

bash

curl -X POST "https://api.cheapaiapi.com/v1m/task/text-to-speech" \
  -H "Authorization: Bearer sk_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Hello, this is a test message.",
    "model": "speech-2.6-hd",
    "voice_setting": {
      "voice_id": "209533299589184",
      "vol": 1,
      "pitch": 0,
      "speed": 1
    },
    "language_boost": "Auto"
  }'

Edge TTS

http

POST https://api.cheapaiapi.com/v1e/task/text-to-speech

textstringrequiredPlain text or valid SRT content.

voicestringVoice name from GET /v1e/voices.

speednumberPlayback speed, for example 0.9 or 1.2.

with_transcriptbooleanGenerate SRT and JSON transcript output.

with_loudnormbooleanNormalize loudness before output delivery.

Task Pattern

json

{
  "success": true,
  "task_id": "uuid_task_id",
  "ec_remain_credits": 123456
}

All current TTS flows are task-based. After submission, poll the common task endpoint or wait for your webhook.

http

GET https://api.cheapaiapi.com/v1/task/{task_id}

Result Payload

json

{
  "id": "task_123",
  "status": "done",
  "credit_cost": 1,
  "metadata": {
    "audio_url": "https://...",
    "srt_url": "https://...",
    "json_url": "https://..."
  },
  "type": "tts"
}

Useful Discovery Endpoints

GET /v1/models for general voice model metadata
GET /v1e/voices for Edge voices
GET /v1m/common/config for MiniMax config and language support
POST /v1m/voice/list for MiniMax voice inventory

Tips For Better UX

Let users choose the TTS route first, then show only the parameters that apply to that route.
Save the last used voice and model combination for quick repeats.
Show task state immediately after submit instead of waiting on audio synchronously.
If transcripts are optional, make them a toggle with clear extra-output expectations.