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

# Voice, Calls, Photos & Files

> Voice messages, AI calls, sending photos/videos/documents

Voice messages, AI calls, sending photos/videos/documents

***

## VOICE MESSAGES AND CALLS

<AccordionGroup>
  <Accordion title="Can the AI recognize voice messages?">
    Yes. The AI automatically transcribes incoming voice messages and replies with text. If needed, you can additionally configure voice replies — see the question "How do I set up the AI voice?".
  </Accordion>

  <Accordion title="How do I set up the AI voice?">
    Configure the voice in a few steps:

    1. Open Channel Advanced Settings → the "Voice" section.
    2. For a standard voice: select one of the preset options and save.
    3. For ElevenLabs Pro voice:

    * Sign up at elevenlabs.io.
    * Choose a ready-made voice or clone your own.
    * Go to your ElevenLabs profile settings → copy the API key.
    * Paste the API key into the corresponding field in the TEXTURA AI settings.
    * Save and test the voice response.
  </Accordion>

  <Accordion title="Can voice responses be disabled?">
    Yes. Go to Channel Advanced Settings → the "Voice" section → deactivate the voice response feature → save.
  </Accordion>

  <Accordion title="Can the AI receive and make calls?">
    Yes, the feature is available. Before connecting, please note the following technical details:

    * Cost: \~\$0.10–0.15 per minute of conversation plus a subscription fee for the number.
    * Inbound calls are generally included in the plan at no extra charge.
    * For Russian numbers there is an integration nuance: SIP telephony is required, and a connection delay is possible.
      To connect, contact your support manager — a specialist will find the optimal configuration for your needs.
  </Accordion>

  <Accordion title="Is it possible to get a Russian number for AI calls?">
    Direct issuance of a Russian number is currently unavailable. Please note the technical specifics:

    * Calls to Russia are possible, but the outgoing number will be registered in another country.
    * When using SIP telephony to route through Russian operators, a characteristic connection delay occurs.
      If this scenario is critical for your business — contact support for a joint analysis of a suitable connection scheme.
  </Accordion>

  <Accordion title="Adjusting voice speed?">
    Go to Advanced Settings → the "Voice" section → the "Speed" parameter. Set a value from x0.5 (slower) to x1.5 (faster) and save.
  </Accordion>
</AccordionGroup>

***

## SENDING PHOTOS, VIDEOS AND FILES

<AccordionGroup>
  <Accordion title="Can the AI send photos and videos?">
    Yes. To configure media file sending:

    1. Upload files to cloud storage (Google Drive, Yandex.Disk, etc.) and obtain direct links.
    2. Add the links to the Talent's instruction or to the "Quick Replies" section.
    3. Specify in which context the AI should send each file.
       Photos and videos are sent as separate messages in the conversation.
  </Accordion>

  <Accordion title="The AI is sending links instead of photos.">
    This is a technical characteristic of the specific data channel:

    * Telegram: in most cases displays the media file directly.
    * Umnico / Bitrix24: the channel may only display a text link — this is an integration nuance on the platform's side.
      For correct display, make sure the link is direct (ends with a file extension: .jpg, .mp4, etc.) and the file is accessible without authentication. If the issue persists, check with support about the behavior of your specific channel.
  </Accordion>

  <Accordion title="How do I send a banquet menu / PDF file?">
    Configure file sending on request:

    1. Upload the PDF to cloud storage (Google Drive, Dropbox, etc.).
    2. Obtain a direct link to the file with public access.
    3. In the Talent's instruction add: "When the menu is requested — send the link: \[your link]".
    4. Save the instruction and test it in a test conversation.
  </Accordion>

  <Accordion title="Can the AI recognize images from guests?">
    Yes. To activate the feature:

    1. Go to Advanced Settings → "Image Recognition" → enable the feature.
    2. In the Talent's instruction describe how the AI should process incoming photos: what to analyze, what to report to the guest based on the result.
    3. Save and test by sending a test image.
  </Accordion>
</AccordionGroup>