https://retellai.com/?utm_source=discord&utm_medium=seo logo
- how can i get Voice Model for voice from API req...
# ❓・ai-evy-support
q
@AI Evy - how can i get Voice Model for voice from API request so i could overwirde agent voice
a
Thanks for tagging me @Q4. I'll look through relevant sources and get back to you shortly.
### Answer You can override an agent's voice model when creating a phone call using the
agent_override
parameter [(1) ]() [(2) ]() . Based on the documentation and community discussions, here's how to structure your request:
Copy code
javascript
const call = await client.call.createPhoneCall({
    to_number: "+1234567890",
    from_number: "+0987654321",
    override_agent_id: "your_agent_id",
    agent_override: {
        agent: {
            voice_id: "minimax-Camille",
            voice_model: "sonic-3"
        }
    }
}) ;
**Important note**: Make sure the
voice_id
and
voice_model
are compatible. For example, don't pair a Cartesia voice with a Minimax model - they need to match their respective TTS providers [(1) ]() . The available voice models include:
eleven_turbo_v2
,
eleven_flash_v2
,
eleven_turbo_v2_5
,
eleven_flash_v2_5
,
eleven_multilingual_v2
,
eleven_v3
,
sonic-3
,
sonic-3-latest
,
tts-1
,
gpt-4o-mini-tts
,
speech-02-turbo
,
speech-2.8-turbo
, and
s1
[(3) ]() . For inbound calls with custom telephony, you can also override the voice settings in your inbound webhook response [(2) ]() .
​ To learn more, see the sources I considered: [[1] Minimax TTS is not outputing audio !](https://discord.com/channels/1186424376300679209/1476976755595018460) [[2] Override agent's configurations before making a call.](https://discord.com/channels/1186424376300679209/1315610729172434945) [[3] Update Voice Agent](https://docs.retellai.com/api-references/update-agent) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.
q
@AI Evy - is there any reported 11labs issues ?
a
Thanks for the follow-up, @Q4. I'll get back to you soon.
### Answer Based on community discussions, there have been reported issues with ElevenLabs voices [(1) ]() [(2) ]() . Voice Model Compatibility Issues [(1) ]() : - When setting
voice_model
to
eleven_turbo_v2_5
via API, some voices are incompatible with this model - Attempting to revert to default models like
eleven_turbo_v2
can result in compatibility errors - The issue is particularly noticeable with OpenAI voices - To reset to default behavior (Auto) , you need to send
null
for
voice_model
, but this must be done in a separate API call before changing the
voice_id
Voice Instability Issues [(2) ]() : - Users reported voices being highly unstable and switching to other voices and languages mid-conversation - Voices randomly speaking in different languages (e.g., French when configured for English) - Pronunciation changing from one call to another - English American accents switching to British accents mid-conversation Suggested Troubleshooting : - Avoid selecting multilingual if not needed - Use GPT-4 or other latest models - Test prompts multiple times - Contact support@retellai.com directly if issues persist These reports come from Discord community discussions and may not represent current platform status. For the most up-to-date information on any ongoing issues, I recommend checking with Retell AI support directly.
​ To learn more, see the sources I considered: [[1] Issue when sending
voice_model
parameter
elevenlabs_turbo_v2_5
](https://discord.com/channels/1186424376300679209/1367836412778512545) [[2] Issues with Elevenlabs?](https://discord.com/channels/1186424376300679209/1363908289133412483) ​
Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.