Analytics preference

We only load Google Analytics after you opt in. You can change this choice at any time from Cookie settings.

Chat Voice (TTS & STT)

The Chat tab can read messages aloud and run a hands-free interactive voice mode, powered by ElevenLabs.

The Chat tab can read messages aloud and run a hands-free interactive voice mode, powered by ElevenLabs. Both directions use one ElevenLabs credential: text-to-speech for playback and Scribe speech-to-text for the microphone.

Setup

  1. Open Settings (gear icon, top right) → Voice tab.
  2. Pick an ElevenLabs credential, or click Add credential to create one. The API key needs the Text to Speech, Speech to Text, and Voices permissions enabled in your ElevenLabs account.
  3. Choose a Voice from your ElevenLabs account.
  4. Click Save Voice Settings.

If you press a voice button before setting this up, the Voice tab opens automatically so you can finish configuring.

The ElevenLabs API key is stored encrypted and never reaches the browser — all calls are proxied through the Heym backend. See Credentials and Settings.

Read a message aloud

Every message bubble (yours and the assistant's) has a speaker button next to Copy. Click it to play the message; click again — or play another message — to stop. Playback uses the ElevenLabs multilingual model, so the language of the text is detected automatically.

Interactive voice mode

The waveform button in the chat input opens a full-screen, hands-free voice mode (light/dark and mobile aware):

  1. It listens through your microphone and detects when you stop speaking.
  2. Your speech is transcribed with Scribe (language auto-detected — no manual TR/EN toggle).
  3. The transcript is sent as a normal chat message in the same conversation.
  4. The assistant's reply is read back aloud, then it resumes listening.

Use the mute button to pause listening and the ✕ to exit (which releases the microphone). Recording works in all modern browsers, including Firefox.

Notes

  • Voice and credential selection are per-user preferences; the active conversation, model, and credential for the LLM are unchanged.
  • ElevenLabs usage (both TTS and STT) is billed to your own ElevenLabs account.
  • Chat Tab – Dashboard chat where voice lives
  • Credentials – How API keys are stored and referenced
  • Settings – Where the Voice tab lives