> ## Documentation Index
> Fetch the complete documentation index at: https://docs.senderz.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Voice

> Text-to-speech voice calls as a flow node and for OTP delivery.

Voice sends a text-to-speech phone call through the same Israeli carrier (019) that powers SMS. It's available two ways: as a **flow node** in an automation, and as an **OTP channel** for one-time codes.

## Voice as a flow node

Add a **Voice call** node to any flow. It reads a text/SMS-format template and speaks the rendered message. The caller ID is your approved sender ID.

The voice worker mirrors the SMS worker gate-for-gate:

<Steps>
  <Step title="Suspension & feature checks">
    A suspended or over-limit workspace can't send.
  </Step>

  <Step title="Sender ID authorisation">
    The caller ID must be an approved sender ID — checked before the wallet.
  </Step>

  <Step title="Suppression, consent, quiet hours">
    Voice rides the **SMS channel** for suppression, `smsConsent`, Israeli
    quiet hours + Shabbat, and frequency caps — a call is at least as
    intrusive as an SMS.
  </Step>

  <Step title="Wallet hold">
    Voice draws down the **SMS credit wallet** at the segment rate, like SMS.
  </Step>
</Steps>

<Warning>
  Marketing voice to Israeli numbers respects quiet hours and Shabbat. Only
  transactional/OTP voice is exempt.
</Warning>

## Voice for OTP

The [OTP API](/en/api-reference/otp) can deliver a one-time code by voice instead of SMS — set `channel: "voice"` on `POST /otp/send`. The code is spoken digit by digit. This is transactional (the recipient requested it), so it skips consent and quiet hours, but still goes through the sender-ID gate and the wallet.

## Delivery details

* Codes are spoken digit-by-digit for clarity.
* Voice is billed at the SMS credit rate (segment count) in v1.
* Consent, suppression and quiet hours use the SMS channel's state — there's no separate voice-consent column in v1.

<Note>
  Voice carries its own send queue and message channel, so voice activity is
  distinguishable in analytics and the audit log even though it shares SMS
  consent and wallet.
</Note>