Skip to content

Ready to build with AnyBot?

AnyBot Docs

Voice calls

Assign a phone number and your agent answers the phone. It greets the caller in a synthesized voice, transcribes what they say, replies out loud, and holds a real spoken conversation.

Give an agent a phone number and it answers incoming calls. On an inbound call the agent picks up, greets the caller in a synthesized voice, listens, transcribes what the caller says, generates a reply using the same knowledge and abilities it uses everywhere else, and speaks the answer back, then keeps the conversation going, turn by turn, until the caller hangs up. The same number also does SMS: one number, both calling and texting.

Before you start

  • A phone number in your space. Get one in the Phone numbers area (search by country or area code; free numbers are available instantly, or paid).
  • An agent assigned to that number. Voice only works once a number has an agent attached. An unassigned number will tell callers to send an SMS instead.
  • Voice conversations are English (see Language below).

Connect

  1. Open Phone numbers and get a number, or pick one you already own.

  2. Assign an agent to the number. The same number then handles both voice and SMS.

  3. Call the number from any phone. The agent answers automatically. No extra voice setup, no Twilio account of your own.

What happens on a call

Caller dials your number
  → Agent answers and greets in a synthesized voice
  → Caller speaks  →  speech is transcribed (English)
  → Agent generates a reply (knowledge · abilities · guardrails)
  → Reply is spoken back
  → (repeat each turn until the caller hangs up)
  1. The agent picks up and opens with a spoken greeting.
  2. It listens and transcribes the caller's speech to text.
  3. It runs that text through the same AI pipeline as chat. Your agent's instructions, knowledge base, and guardrails all apply.
  4. It synthesizes the reply to speech and plays it back.
  5. It listens for the next turn and repeats. After a short silence with nothing spoken, the agent says goodbye and ends the call.

On a call

CapabilityDetail
Answer inbound callsAgent auto-answers any call to an assigned number
Speech-to-textCaller audio is transcribed each turn
Spoken repliesReplies are synthesized and played back in a natural voice
Multi-turnBack-and-forth continues until the caller hangs up
TranscriptEvery caller turn and agent reply is saved to the conversation
Call logDirection, from/to number, status, and duration are recorded

Language

Voice is English-only. Caller speech is transcribed as English and the agent's spoken replies are English. This is a fixed limit of the voice channel today. It's independent of your agent's language settings on text channels. If a caller needs another language, route them to a text channel like WhatsApp or web chat.

Transcripts and recordings

  • Transcripts are real. Each call becomes a conversation: the transcribed caller turns and the agent's spoken replies are stored as messages, so a voice call reads back in your inbox exactly like a chat thread. Every call is also logged with its direction, numbers, status, and duration.
  • Audio recordings are off by default. The platform can store a recording URL when call recording is enabled on the number, but the standard voice flow doesn't record audio. Don't assume calls are recorded unless your operator has turned recording on.

Channel settings

Voice uses the same agent as every other channel. Its instructions, knowledge, and abilities carry over unchanged. See Channel settings for the shared controls, and Phone numbers for buying, assigning, and releasing numbers.

Troubleshooting

  • Caller is told to "send an SMS instead": the number has no agent assigned. Assign an agent to it in Phone numbers.
  • "This number is not available": the number isn't wired to your space. Re-check it in Phone numbers.
  • Agent didn't catch the caller: if the line is silent or too noisy for transcription, the agent says it didn't catch that and ends the call (there's no second attempt). Ask the caller to speak clearly and call back.
  • No voice option in the space: the channel hasn't been enabled for your plan; contact your operator.

FAQ

Can the AI really answer a phone call? Yes. It picks up, understands the caller, replies out loud, and keeps talking across turns.

Do I need my own Twilio account? No, get a number inside AnyBot and assign an agent.

Same number as SMS? Yes, one number does both SMS and voice.

What languages does it speak? English only, on voice.

Is the call recorded? You always get the transcript. Audio recording is only captured if your operator enables it.

Next

  • Phone numbers: search, buy, and assign the number voice runs on.
  • SMS: the texting side of the same number.
  • Channel settings: the shared controls that apply across channels.
  • All channels: connect WhatsApp, Telegram, web chat, and more.