Skip to main content
Configure how your agent speaks using the synthesizer settings.

​
Synthesizer Settings

The synthesizer object controls text-to-speech: 
synthesizer = {
    "voiceConfig": {
        "model": "waves_lightning_large",
        "voiceId": "zorin",
        "gender": "male"
    },
    "speed": 1.2,
    "sampleRate": 24000
}

​
Voice Config

FieldTypeDescription
modelstringTTS model to use
voiceIdstringVoice identifier
genderstring”male” or “female”

​
Available Models

ModelDescription
wavesStandard quality
waves_lightning_largeHigh quality, fast
waves_lightning_v2Latest version
otherExternal provider (e.g., ElevenLabs)

​
Speed

Control speaking rate: 
"speed": 1.0   # Normal (default)
"speed": 1.2   # 20% faster
"speed": 0.8   # 20% slower
Range: 0.5 to 2.0

​
Sample Rate

Audio quality setting: 
"sampleRate": 24000  # Standard quality
"sampleRate": 44100  # High quality

​
Example: Configure Agent Voice

import requests
import os

API_KEY = os.getenv("SMALLEST_API_KEY")
agent_id = "your-agent-id"

# Update agent with custom voice
response = requests.patch(
    f"https://atoms.smallest.ai/api/v1/agent/{agent_id}",
    headers={
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    },
    json={
        "synthesizer": {
            "voiceConfig": {
                "model": "waves_lightning_large",
                "voiceId": "nyah",
                "gender": "female"
            },
            "speed": 1.1,
            "sampleRate": 24000
        }
    }
)

print(f"Updated: {response.status_code}")

​
Background Sound

Add ambient audio: 
"backgroundSound": "call_center"  # Call center ambience
"backgroundSound": "office"       # Office sounds
"backgroundSound": "cafe"         # Cafe background
"backgroundSound": ""             # No background (default)