Skip to main content
Voice lets your agents talk to callers over the phone. You create an agent, set how it speaks and listens, and connect it to a phone number. Cognigy handles speech recognition, conversation, and speech synthesis during the call.

Get started

1

Check your setup

Review what voice supports and check that your project is ready:

Explore technical specifications

Review the supported protocols, audio codecs, trunk routing, speech providers, call recording, and call quality tools.

Configure speech and telephony

Ask your administrator to configure the speech providers and telephony connections for your project.
2

Choose how to build your voice agent

Decide whether to use agents or flows. Select a tab to see the voice features for each.
The following features are available only for agents:

Voice studio

Set the speech providers, voice, greeting, turn taking, and listening options for an agent. Not available when speech to speech is on.

Voice tools

Let an agent hang up, transfer, or play audio during a call.

Speech to speech

Experimental. Run a real-time voice model for natural conversation, backed by your existing agent. Replaces the text-to-speech and speech-to-text settings.

Try additional features

The following features work with agents and flows:

Outbound calls

Start calls from Cognigy to a phone number or another system, and track call statuses through webhooks and call events.

TTS streaming

Play synthesized speech as it’s generated to reduce response latency.

Click To Call

Embed a voice widget in your website so that customers can speak directly with agents.

Voice endpoint

Connect an agent to your contact center or other phone numbers with the Voice endpoint.

More information

Last modified on November 11, 2026