tts-production

TTS Production with ElevenLabs

Generate, tune, and mix voice-over audio using the ElevenLabs Text-to-Speech API.

Critical: Use direct API calls (curl), NOT ElevenLabs MCP tools. The MCP Text_To_Speech tool returns 404 due to routing issues. Direct API is reliable and battle-tested.

API Pattern

Text-to-Speech with Timestamps (recommended)

curl -s -X POST "https://api.elevenlabs.io/v1/text-to-speech/${VOICE_ID}/with-timestamps" \
  -H "xi-api-key: ${ELEVENLABS_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Your text here",
    "model_id": "eleven_multilingual_v2",
    "voice_settings": {
      "stability": 0.75,
      "similarity_boost": 0.80,
      "style": 0.40,
      "use_speaker_boost": true
    }
  }' \
  --output /tmp/tts-response.json

Decode Response

import json, base64
with open('/tmp/tts-response.json', 'r') as f:
    data = json.load(f)
audio_bytes = base64.b64decode(data['audio_base64'])
with open('output.mp3', 'wb') as f:
    f.write(audio_bytes)
ends = data.get('alignment', {}).get('character_end_times_seconds', [])
print(f'Duration: {ends[-1]:.2f}s' if ends else 'No timestamps')

Simple TTS (no timestamps)

curl -s -X POST "https://api.elevenlabs.io/v1/text-to-speech/${VOICE_ID}" \
  -H "xi-api-key: ${ELEVENLABS_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{ "text": "...", "model_id": "eleven_multilingual_v2", "voice_settings": {...} }' \
  --output output.mp3

Sound Effects Generation

curl -s -X POST "https://api.elevenlabs.io/v1/sound-generation" \
  -H "xi-api-key: ${ELEVENLABS_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{ "text": "short sharp underwater splash blip", "duration_seconds": 1.0, "prompt_influence": 0.8 }' \
  --output sfx.mp3

Model Selection

Model	Use Case	Speed	Quality
eleven_multilingual_v2	Production — Dutch, English, mixed language	Slow	Highest
eleven_flash_v2_5	Quick drafts, iteration	Fast	Good
eleven_turbo_v2_5	Real-time, low latency	Fastest	Acceptable