### Answer --- Yes, Retell AI offers voicemail detection that you can enable on your agent. [(1) ]() [(2) ]() Voicemail Detection Settings: -

enable_voicemail_detection

— Enable this to detect voicemails. [(2) ]() -

voicemail_message

— The message to play when voicemail is detected (set to empty string to hang up instead) . -

voicemail_detection_timeout_ms

— Configures when to stop running voicemail detection (default 30,000 ms, range 5,000–180,000 ms) . -

voicemail_option

— Defines the action to take when voicemail is detected (e.g., hang up or leave a message) . Filtering in Call History: For filtering calls that hit voicemail or were silent, you can use custom filters in Call & Chat History. The changelog notes that call history now supports filtering by custom post-call analysis fields, including enum, boolean, and number types. [(3) ]() For batch calls where some users didn't answer, you can filter failed/unanswered calls from the call history, export them to CSV, and re-upload them to the batch dialer. [(4) ]() ---

​ To learn more, see the sources I considered: [[1] Update Voice Agent](https://docs.retellai.com/api-references/update-agent) [[2] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) [[3] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) [[4] Batch call](https://discord.com/channels/1186424376300679209/1357080050343743508) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

@AI Evy so where do I put the VM detection settings?

### Answer Voicemail detection is configured at the agent level under Call Settings in the dashboard. [(1) ]() Steps: 1. Navigate to your agent settings → Call Settings 2. Enable "Voicemail Detection" 3. Choose the voicemail behavior: - **Hang up**: The agent disconnects when voicemail is detected - **Leave a message**: The agent leaves a message (static text or prompt-generated) Via API: You can also set these options when creating or updating an agent using the API: [(2) ]() [(3) ]()

javascript
enable_voicemail_detection: true,
voicemail_message: "Hi, please give us a callback.",
voicemail_detection_timeout_ms: 30000,
voicemail_option: {
  action: {
    type: "static_text",
    text: "Please give us a callback tomorrow at 10am."
  }
}

[(2) ]() If a call is detected as voicemail and the configured action is taken successfully, the call will have a disconnection reason of

voicemail_reached

. [(1) ]()

​ To learn more, see the sources I considered: [[1] Handle voicemail](https://docs.retellai.com/build/handle-voicemail) [[2] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) [[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.

​ To learn more, see the sources I considered: [[1] Update Voice Agent](https://docs.retellai.com/api-references/update-agent) ​

### Answer ### Answer I don't have explicit documentation on whether agent setting changes apply in real-time to an ongoing batch campaign or only to new calls. Based on the available sources, you can update agent settings via the API using the Update Voice Agent endpoint. [(1) ]() However, the documentation doesn't specify whether those changes affect calls already queued in a batch dialer or only apply to newly initiated calls. Recommendation: To be safe, if you need settings changes to apply to all remaining calls, you may want to pause/stop the current batch and restart with the updated agent configuration. Check out these potentially useful sources: - [Update Voice Agent]() - [Platform Changelogs | Retell AI]() - [Step 1: Configure global settings]() Would you like me to try to answer generally based on my knowledge instead?

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer --- There are several voice settings you can adjust to improve how natural the agent sounds: [(1) ]() [(2) ]() [(3) ]() Voice Temperature (

voice_temperature

) : Controls voice stability. Value ranges from 0–2. Lower = more stable, higher = more variant/expressive speech. Currently only applies to 11labs voices. Default is 1. [(1) ]() Voice Speed (

voice_speed

) : Controls speech rate. Value ranges from 0.5–2. Lower = slower, higher = faster. Default is 1. Voice Model (

voice_model

) : Different voice models may produce different quality. Options include

eleven_turbo_v2

,

eleven_flash_v2

,

eleven_turbo_v2_5

,

eleven_flash_v2_5

,

eleven_multilingual_v2

,

sonic-2

,

sonic-3

,

sonic-turbo

,

tts-1

,

gpt-4o-mini-tts

,

speech-02-turbo

. Normalize for Speech (

normalize_for_speech

) : If set to true, normalizes numbers, currency, dates, etc. to spoken form for more consistent speech synthesis. [(2) ]() Things to try: - Increase

voice_temperature

slightly (e.g., 1.2–1.5) for more natural variation [(1) ]() - Try a different

voice_model

if available for your voice provider - Check if

normalize_for_speech

is enabled to help with number/date pronunciation [(2) ]() You can update these via the [Update Voice Agent]() API or in the dashboard. [(3) ]()

​ To learn more, see the sources I considered: [[1] Get Voice Agent](https://docs.retellai.com/api-references/get-agent) [[2] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) [[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.

@AI Evy What's the best settings for the natural then

### Answer I don't have documentation specifying "best" or recommended values for the most natural-sounding voice. The sources describe what the settings do and their valid ranges, but don't provide optimal configurations. [(1) ]() [(2) ]() What you can adjust: - `voice_temperature`: Range [0, 2]. Lower = more stable, higher = more expressive/variant. Default is 1. Only applies to 11labs voices. [(1) ]() - `voice_speed`: Range [0.5, 2]. Lower = slower, higher = faster. Default is 1. - `voice_model`: Options include

eleven_turbo_v2

,

eleven_turbo_v2_5

,

eleven_multilingual_v2

,

sonic-2

,

sonic-3

, etc. [(2) ]() - `normalize_for_speech`: When

true

, converts numbers/dates/currency to spoken form for more consistent synthesis. Suggestions to experiment with: - Try slightly increasing

voice_temperature

(e.g., 1.1–1.3) for more natural variation [(1) ]() - Test different

voice_model

options to see which sounds best for your use case [(2) ]() - Enable

normalize_for_speech

if you have numbers or dates in your responses Since optimal settings depend on your specific voice and use case, experimentation is recommended.

​ To learn more, see the sources I considered: [[1] List Voice Agents](https://docs.retellai.com/api-references/list-agents) [[2] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.