hello | ElevenLabs | Page 1

hazy kestrelBOT Aug 21, 2025, 11:25 AM

#

AI-generated Answer for baymurat1

Hello! How can I help you with ElevenLabs today? Tell me what you’re trying to do, and I’ll pull the exact steps from the official docs with citations.

unkempt mason Aug 21, 2025, 11:25 AM

#

I am having this issue: This conversation failed with the following reason:

Override for field 'voice_id' is not allowed by config.

hazy kestrelBOT Aug 21, 2025, 11:25 AM

#

AI-generated Answer for baymurat1

This happens when you pass a conversation override for tts.voice_id, but your agent’s Security settings don’t allow that field to be overridden. Overrides are disabled by default, and the platform throws an error if you provide an override for a field that isn’t enabled. (elevenlabs.io)

Fix

In the ElevenLabs dashboard, open your agent > Settings > Security, and enable the Voice ID override. Then your tts.voice_id override will be accepted. (elevenlabs.io)
Or, remove tts.voice_id from conversation_config_override so the agent uses its default voice. (elevenlabs.io)

Reference structure for passing an override (SDK):

conversation_config_override: { tts: { voice_id: "your_voice_id" } } — only after enabling the Voice ID override in Security. (elevenlabs.io)

Sources:
Overrides | ElevenLabs Documentation
Overrides | ElevenLabs Documentation
Overrides | ElevenLabs Documentation
Overrides | ElevenLabs Documentation

unkempt mason Aug 21, 2025, 11:37 AM

#

can I edit voice speed in conversation ai?

hazy kestrelBOT Aug 21, 2025, 11:37 AM

#

AI-generated Answer for baymurat1

Yes. Conversational AI supports a “Speed” setting for each voice. You can set it between 0.7 and 1.2 (default 1.0); values below 1.0 slow speech and values above 1.0 speed it up. Extreme values may reduce audio quality. (elevenlabs.io, help.elevenlabs.io)

If you’re configuring via API/SDK, set the speed field in the voice settings (e.g., POST /v1/voices/:voice_id/settings/edit with "speed": 0.9). (elevenlabs.io)

Sources:
Speed control | ElevenLabs Documentation
Can I change the pace of the voice? – ElevenLabs
Edit voice settings | ElevenLabs Documentation

unkempt mason Aug 21, 2025, 11:51 AM

#

I don't see any change in my speed

hazy kestrelBOT Aug 21, 2025, 11:51 AM

#

AI-generated Answer for baymurat1

Thanks for the details. Speed control is supported in Conversational AI, but a few common misconfigurations can make it seem like nothing changed.

What to check

Set speed on the agent’s TTS settings: In your agent, set speed to a value between 0.7 and 1.2 (default is 1.0). Values below 1.0 slow speech; above 1.0 speed it up. Try a bigger change (e.g., 0.85 or 1.15) to make the difference obvious.
If you use the API/SDK, set tts.speed on the agent: The agent config supports a speed property under conversation_config.tts.
Using Multi‑voice? A per‑voice speed can override the base agent speed. Check the “Supported voices” section for the active voice and its parameters.
Don’t pass unsupported values: Only 0.7–1.2 are supported; values outside this range won’t work and may degrade quality near the extremes.

About your earlier “Override for field 'voice_id' is not allowed by config.”

If you’re trying to change voice (or language/prompt) at runtime via widget attributes or ConversationInitiationData, you must enable the corresponding override in the agent’s Security tab first. Otherwise the platform returns an override-not-allowed error.

Optional: voice-level defaults

You can also read/update a voice’s own speed setting via the Voices API, but for Conversational AI you typically set speed on the agent.

If you’ve done the above and still hear no change, tell me how you’re configuring speed (dashboard vs. API), whether Multi‑voice is enabled, and the exact value you set—I’ll help pinpoint the issue.

unkempt mason Aug 21, 2025, 11:55 AM

#

should I enable this?

Fetch initiation client data from webhook

If enabled, the conversation initiation client data will be fetched from the webhook defined in the settings when receiving Twilio or SIP trunk calls.

hazy kestrelBOT Aug 21, 2025, 11:55 AM

#

AI-generated Answer for baymurat1

Short answer: enable it only if you want to personalize inbound Twilio or SIP trunk calls by pulling per‑call data from your own webhook. Otherwise, leave it off.

When to enable

You have configured a webhook URL in Settings and need to pass dynamic_variables and/or conversation_config_override at call start (e.g., personalize prompts or switch voice_id per caller). (elevenlabs.io)
For Twilio inbound calls, ElevenLabs will POST caller_id, agent_id, called_number, and call_sid to your webhook and use the JSON it returns to initialize the conversation. Your webhook must respond quickly to avoid call delays. (elevenlabs.io)
For SIP trunk calls, caller/call IDs from headers are exposed to the pre‑call webhook and as system dynamic variables (supports X-CALL-ID/X-CALLER-ID and Twilio’s sip.twilio.callSid fallback). (elevenlabs.io)

Important for your earlier error

The error “Override for field 'voice_id' is not allowed by config.” occurs if you return an override for a field that isn’t permitted in the agent’s Security tab. Only these override fields are supported: System prompt, First message, Language, Voice ID. Enable the specific fields you plan to override before using the webhook. (elevenlabs.io)

Note on speed

Speaking speed is controlled via the agent’s voice settings (speed 0.7–1.2). It’s not one of the supported override fields, so set it in the voice configuration rather than via webhook. (elevenlabs.io)

unkempt mason Aug 21, 2025, 11:57 AM

#

If you’ve done the above and still hear no change, tell me how you’re configuring speed (dashboard vs. API), whether Multi‑voice is enabled, and the exact value you set—I’ll help pinpoint the issue.

Look, in dashaboard I have set the speed to 1, but I want to overwrite it by setting 0.8 for one voice id. can i do it?

hazy kestrelBOT Aug 21, 2025, 11:57 AM

#

AI-generated Answer for baymurat1

Short answer: Yes—set speed per voice, but not via per‑conversation overrides.

Ways to do it

Edit that voice’s settings to speed 0.8. This applies wherever that voice_id is used. You can do this in the Voice settings UI or via the “Edit voice settings” API (body includes speed; valid range 0.7–1.2, default 1.0). (elevenlabs.io)
If you use Multi‑voice, you can configure per‑voice TTS settings (including speed) for each “Supported voice” inside the agent. Set that specific supported voice to 0.8 so it only affects that agent/label. (elevenlabs.io)

What you cannot do

You cannot pass speed through conversation overrides; overrides support agent prompt, first_message, language, and tts.voice_id only. Attempting to override a field that isn’t enabled (or isn’t supported) throws an error like the one you saw. If you need to override voice_id at start, enable it in the agent’s Security > Overrides first. (elevenlabs.io)

Reference on speed behavior and limits: values below 1.0 slow speech, above 1.0 speed it up; supported range is 0.7–1.2. (elevenlabs.io)

If you want, share whether Multi‑voice is enabled and which path you prefer (voice settings vs. Multi‑voice per‑voice config), and I’ll outline the exact clicks or API call. (elevenlabs.io)

unkempt mason Aug 21, 2025, 12:06 PM

#

so I cannot change the voice speed for a specific voice id

hazy kestrelBOT Aug 21, 2025, 12:06 PM

#

AI-generated Answer for baymurat1

Short answer: you can.

What works

Per-voice setting (recommended): Set speed on that specific voice_id so it applies wherever that voice is used, including Conversational AI. Use PATCH /v1/voices/:voice_id/settings with {"speed": 0.8}, or change it in the voice’s settings in the dashboard. You can read or verify with GET /v1/voices/:voice_id/settings. (elevenlabs.io)
Per‑request (TTS APIs): You can also override speed at request time via voice_settings.speed in HTTP or WebSocket TTS calls. (elevenlabs.io)
Product support and range: Speed control exists in Conversational AI, with a supported range of 0.7–1.2 (1.0 default). (elevenlabs.io, help.elevenlabs.io)

What doesn’t work the way you tried

Agent overrides: In Conversational AI, the agent-level “Overrides” feature allows overriding Voice ID (and some other fields) at call start, but it doesn’t document speed as an overridable field. If you saw “Override for field 'voice_id' is not allowed by config,” enable the Voice ID override in the agent’s Security tab first. To change only one voice’s speed, set it on that voice_id (as above) rather than relying on agent overrides. (elevenlabs.io, help.elevenlabs.io)

Summary

Yes—you can run your agent at Speed=1 globally, and set Speed=0.8 only for a particular voice by editing that voice’s settings (API or dashboard). This is the supported, per‑voice way to do it in Conversational AI. (elevenlabs.io)

Sources:
Edit voice settings | ElevenLabs Documentation
Create speech | ElevenLabs Documentation
Speed control | ElevenLabs Documentation
Can I change the pace of the voice? – ElevenLabs
Overrides | ElevenLabs Documentation
Where can I locate the reason for my call failing? – ElevenLabs
Edit voice settings | ElevenLabs Documentation

unkempt mason Aug 21, 2025, 12:10 PM

#

can I use this for Conversation AI?

Edit voice settings
POST

https://api.elevenlabs.io
/v1/voices/:voice_id/settings/edit
POST
/v1/voices/:voice_id/settings/edit

cURL

curl -X POST https://api.elevenlabs.io/v1/voices/voice_id/settings/edit
-H "xi-api-key: xi-api-key"
-H "Content-Type: application/json"
-d '{
"stability": 1,
"use_speaker_boost": true,
"similarity_boost": 1,
"style": 0,
"speed": 1
}'
Try it
200
Successful

{
"status": "ok"
}
Edit your settings for a specific voice. “similarity_boost” corresponds to “Clarity + Similarity Enhancement” in the web app and “stability” corresponds to “Stability” slider in the web app.

Path parameters
voice_id
string
Required
ID of the voice to be used. You can use the Get voices endpoint list all the available voices.

Headers
xi-api-key
string
Required
Request
This endpoint expects an object.
stability
double or null
Optional
Determines how stable the voice is and the randomness between each generation. Lower values introduce broader emotional range for the voice. Higher values can result in a monotonous voice with limited emotion.
use_speaker_boost
boolean or null
Optional
This setting boosts the similarity to the original speaker. Using this setting requires a slightly higher computational load, which in turn increases latency.
similarity_boost
double or null
Optional
Determines how closely the AI should adhere to the original voice when attempting to replicate it.
style
double or null
Optional
Determines the style exaggeration of the voice. This setting attempts to amplify the style of the original speaker. It does consume additional computational resources and might increase latency if set to anything other than 0.
speed
double or null
Optional
Adjusts the speed of the voice. A value of 1.0 is the default speed, while values less than 1.0 slow down the speech, and values greater than 1.0 speed it up.

#hello