Text-to-Speech

Generate speech audio from text.

Endpoint

POST /v2/audio/tts/sesame

Request

{
  "text": "Hello world",
  "preset_voice": "Alice"
}

Parameter	Type	Description
text	string	Text to speak (required, max 10000 chars)
preset_voice	string	Voice preset (see below)
custom_voice	object	Custom voice cloning (see below)

Response

Returns audio/wav binary data.

curl -X POST https://relay.opengpu.network/v2/audio/tts/sesame \
  -H "X-API-Key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "Hello world", "preset_voice": "Alice"}' \
  -o output.wav

Voice Presets

curl https://relay.opengpu.network/v2/audio/tts/presets \
  -H "X-API-Key: YOUR_KEY"

Available: Alice, Avery, Brock, Chloe, Ella, Emma, Grace, Karen, Kevin, Lucas, Matt, William

Custom Voice

Clone any voice with a sample audio URL and optional context text:

{
  "text": "Hello in my custom voice",
  "custom_voice": {
    "url": "https://example.com/voice-sample.wav",
    "context_text": "Text spoken in the reference audio"
  }
}