Documentation

Introduction

Press or Say - AI Voice Recognition for IVR Menus

View as Markdown

Callers can press a key or speak their choice. Each delivered speech-enabled Gather is billed - including silent or keypad-only responses and retries.

Enabling the feature

  1. Open the prompt (Greeting, zip, tag question, etc.).
  2. Check Enable voice recognition. It works with TTS and Audio files that are at most 55 seconds long. Screenshot 2026 10 05 at 13.45.15
  3. Set up the routing options depending on what topic is extracted from the caller’s response: Screenshot 2026 10 05 at 13.46.25

Verifying it works

You will see logs in your call quoting the caller’s response and how the call was routed based on that response: ![Screenshot 2026Screenshot 2026 10 05 at 13.58.08

Constraints that matter (read before flipping it on)

Billing

Each speach-enabled Gather is charged per successfully delivered attempt.

Undelivered gathers aren’t billed. Charges show up with your other call-related line items (memo: IVR Voice Recognition).

Contact your sales representitive for an exact quote.

Help us improve this article or request new support guides.