Getting Started

Braiv Speech is an ElevenLabs-compatible Text to Speech API for existing Braiv voice clones. It is a sibling of Public API v1 (video ingest, dubs, jobs, webhooks). Scoped braiv_live_… keys work on Public API; legacy unscoped keys remain TTS-only.

Base URL

https://api.braiv.co/v1

Requirements

• A Braiv account on a Pro, Business, or Team plan.
• An API key generated from the Braiv app.
• A Braiv voice clone voice_id. Create one with POST /v1/voices/add, then poll GET /v1/voices/{voice_id} until status is ready. Studio-created clones work the same way.

First request

curl -X POST "https://api.braiv.co/v1/text-to-speech/YOUR_VOICE_ID" \
  -H "xi-api-key: YOUR_BRAIV_API_KEY" \
  -H "Content-Type: application/json" \
  --output speech.wav \
  --data '{
    "text": "The first move is what sets everything in motion.",
    "model_id": "braiv-speech-v1",
    "language_code": "en",
    "voice_settings": {
      "speed": 1
    }
  }'

The beta returns WAV audio. MP3, PCM variants, and streaming are planned future additions.

OAuth Connect tokens (oat_) work on Speech as well as Public API, so hosted MCP can list, clone, and speak with the same token. See Voices.

For video workflows (import, dub, shorts, packaging), use the Public API.