ghostdragon
03/24/2025, 5:39 PMVapi Support Bot
03/24/2025, 5:41 PMTo enable SSML in VAPI, log in, navigate to voice settings, enable SSML parsing, and save changes.
[According to this Discord thread](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1296031572956549171), you can select your current voice provider (like 11labs) and enable enableSsmlParsing
.
### 2. Adjust Speech Rate Settings
Look for parameters in your voice synthesis software like rate
, speed
, or tempo
and set them to lower values to slow down speech. [This Discord discussion](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1336023375721791549) mentions these parameters can help control speech rate.
### 3. Configure Chunk Plan Settings
Tune the chunkPlan
configuration to control how text is processed:
json
{
"voice": {
"chunkPlan": {
"enabled": true,
"minCharacters": 30, // Try adjusting between 1-80
"punctuationBoundaries": [".", "!", "?", ";"]
}
}
}
[According to this Discord thread](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1343543475383046185), keeping chunkPlan.enabled: true
is important for proper sentence boundary detection.
### 4. Adjust Stop Speaking Plan
The stopSpeakingPlan
configuration can help optimize the timing of speech:
numWords: 0
voiceSeconds: 0.2
backoffSeconds: 1
[This is mentioned in Vapi's documentation](https://docs.vapi.ai/customization/speech-configuration) for controlling speech timing.
### 5. Add Hesitations and FillersVapi Support Bot
03/24/2025, 5:41 PMKyle Brunker
03/26/2025, 1:16 PM