Skip to main content

Voice Management

Streamerly provides a library of voices across all TTS providers. You can select specific voices for your triggers, let the system randomize between them, or allow viewers to choose their own.

Voice Properties

Each voice has the following attributes:

PropertyDescription
Display NameHuman-readable voice name
VendorTTS provider (ElevenLabs, Azure, Amazon, DECTalk)
EngineEngine type (standard, neural, AI, DECTalk)
GenderMale, female, or unknown
LanguageBCP 47 language code (e.g., en-US, ja-JP)

Selecting Voices

In your trigger's TTS settings:

  1. Default Voice — Set the primary voice used for all alerts
  2. Voice List — Add multiple voices to the available pool
  3. Randomized — If enabled, a random voice from the list is used each time
  4. User Choice — If enabled (for Super Cheers), viewers can pick from the list

Voice Cloning (ElevenLabs)

ElevenLabs supports custom voice cloning:

  • Cloned voices are private to the user who created them
  • They appear alongside public voices in your voice selection
  • Cloned voices are identified by an is_cloned flag and tied to their owner

DECTalk Voices

DECTalk offers 6 built-in voices with distinctive retro characteristics:

VoiceDescription
PaulDefault male voice
BettyFemale voice
HarryDeep male voice
FrankOlder male voice
DennisNasal male voice
KitChild voice

Tips

  • Preview voices before selecting them — each voice has a distinct sound
  • Match voice to content — Natural AI voices work well for serious alerts, DECTalk is great for fun/meme content
  • Use randomization for variety — viewers enjoy hearing different voices
  • Filter by language if your audience speaks a specific language

See Also