Audio settings control the platform's listening and speaking behaviour: dictation into chats, and spoken playback of answers.
Options
| Option | Effect |
|---|---|
| Voice | The voice used to read responses aloud. The list contains the voices enabled in your deployment — if it's empty ("No voices available"), text-to-speech isn't enabled in your workspace. |
| Speech Rate | Reading speed — slower for presentations and accessibility, faster for scanning. |
| Transcription Language | The language spoken dictation is transcribed in. Auto-detect works for clear single-language speech; picking your language explicitly improves accuracy, especially with accents or jargon. |
| Auto-Playback Response | Speaks each new answer automatically — useful for hands-free or screen-limited work. |
| Reset | Restores your audio preferences to defaults. |
| System default | Lets your device/browser audio configuration apply. |
Using dictation
- Place the cursor in any message box.
- Select the microphone control and speak.
- Review the transcription (edit if needed) and send.
Where configured, messages can send automatically once transcription finishes. See Chat Basics.
Tips
- Name recognition: people's names, product codes and acronyms transcribe better when you dictate them in chunks or type the odd token afterwards.
- Quiet rooms win: noise hurts accuracy far more than accent does.
- Switch languages explicitly when you work in more than one — Auto-detect can mis-follow language changes mid-sentence.
- If microphone access fails, check your browser's permission for the site.
Availability
Voice features depend on administrator enablement (Audio (Admin)). Voice-mode conversations, when enabled, let you talk to the assistant hands-free.