Back to all skills
๐Ÿ”Š
Communication

Voice Reply

Have the assistant reply in voice โ€” natural, low-latency TTS.

4.7rating
12,400 installs
hermesrecall/voice-reply
Pro required

About this skill

Voice Reply is a native HermesRecall skill that wraps OpenAI's TTS API to produce a clean, natural voice answer. Pair it with Voice Transcription and the loop is fully conversational: speak, listen, speak again. Picks the right voice and locale automatically, caches repeated phrases, and streams audio for sub-second time-to-first-sound.

What it does

  • Natural TTS voices (FR + EN + more)
  • Streaming playback for fast response
  • Auto voice selection based on user locale
  • Phrase caching for common replies
  • Plays via system audio or returns an mp3 blob

Use cases

  • Get spoken answers while cooking or driving
  • Voice-reply to a WhatsApp message you just dictated
  • Generate audio briefings for tomorrow morning