Skip to content
Docs
Voxbi Cockpit APIs
API Guides

Announcements

An announcement is a piece of audio your PBX plays to callers: an IVR greeting, a closed-hours message, a queue context line. Create one once and point as many things at it as you like, because the audio is independent of everything referencing it.

Overview

Create an announcement once and reference it from any number of places: IVRs use them as the menu prompt, call flow Play announcement steps play them mid-call, and queues use them as positional announcements. The audio is independent of its references, so editing or replacing the file updates everywhere it is used.

Source What it is When to use
Uploaded file A pre-recorded MP3, WAV, FLAC, or AAC file up to 20 MB Professional recordings, branded greetings, anything where voice nuance matters
Text to speech TTS generated from a written message and a chosen voice Frequently-changing announcements, multi-language variants, quick iteration

Files are converted internally to telephony-friendly codecs, so you do not need to pre-convert them.

What it does

The Source field shapes everything else, and switching source does not preserve fields from the previous one.

When Source is Text to speech, a three-step panel appears: select a voice (filtered by language, gender, accent, use case), write your message (up to 5,000 characters, previewable aloud), and adjust four voice settings before generating:

  • Stability (0-1): lower is more expressive, higher more consistent. Default 0.5.
  • Similarity boost (0-1): how closely to match the trained voice. Default 0.75.
  • Style (0-1): how much character to inject. Default 0.
  • Speed (0.7x-1.2x): playback tempo. Default 1.0.

Generate to synthesize the audio. Editing a TTS announcement later keeps the message and settings so you can iterate.

See also

  • IVRs - the most common consumer of announcements.
  • Call flows - the Play announcement and Text to speech steps that play announcements mid-call.
  • Music on hold - separate but related: audio for callers waiting in queues.
  • Sound files - the dashboard that groups all three audio types.

Form fields

Fields available in the create and edit forms.

General

Name

Identifies this announcement in dropdowns and lists. Use a name that describes what the audio says, not where it is used.

Comment

Internal note for your own reference, such as the source recording or script version.

Language

The language spoken in this announcement. Optional for a plain announcement, but required once it has language variants (and on every variant). When the PBX detects the caller's language, it plays the matching variant instead of this announcement; otherwise this one is the fallback.

Source

How the audio was created. Uploaded file for a manually recorded file, or Text to speech to generate it from a written message.

  • uploaded_file Uploaded file Upload an MP3, WAV, FLAC, or AAC file up to 20 MB. The system converts it internally.
  • ai_generated Text to speech Generate audio from a written message and a chosen voice. Reveals a three-step generation panel.
AI text-to-speech

Type your message, pick a voice, and generate the file. Only available when Source is set to Text to speech.

Sound file

The audio file to upload. Accepted formats are MP3, WAV, FLAC, and AAC up to 20 MB.

Current audio

Plays back the currently saved audio so you can preview it without downloading.

Updated