Voice Interaction¶
Talk to Tessa like a teammate. Ask questions out loud, get spoken answers back - no keyboard required.
Voice interaction lets you have a natural conversation with Tessa using speech recognition and synthesis. This is especially valuable in VR, where typing is impractical and hands-free operation keeps you immersed in the 3D environment.
Activating Voice¶
- Click the microphone icon in the chat panel to begin listening
- Type
/voice onto enable voice mode - Type
/voiceto toggle voice on or off
Once active, Tessa listens for your question, processes it, and responds with synthesized speech while also showing the text response in the chat panel.
Voice Commands¶
| Command | What It Does |
|---|---|
/voice | Toggle voice mode on/off |
/voice on | Enable voice mode |
/voice off | Disable voice mode |
/mute | Mute Tessa's speech output (she still responds in text) |
/unmute | Re-enable Tessa's speech output |
Mute vs voice off
/mute silences Tessa's spoken responses but keeps voice recognition active - you can still speak to her. /voice off disables both speech input and output.
Changing Tessa's Voice¶
Tessa can speak in different voices. Use the voice tool to browse and switch:
| What to Say | Effect |
|---|---|
| "List available voices" | Shows voices filtered by locale and gender |
| "Preview the Jenny voice" | Plays a sample of a specific voice |
| "Change your voice to Jenny" | Switches to a named voice |
| "What voice are you using?" | Shows current voice |
Tessa supports fuzzy matching - saying "Jenny" resolves to the full voice name (e.g., "en-US-JennyNeural"). Your voice selection persists across sessions.
Interrupting Tessa¶
If Tessa is speaking and you need to cut in:
- Press Escape to stop her current speech immediately
Pressing Escape does not lose your place in the conversation. Tessa's text response remains in the chat, and you can continue asking follow-up questions right away.
Tips for VR Headset Users¶
Voice is the primary interaction mode for VR:
- Keep voice mode on so the microphone stays active between questions. You can have a continuous conversation without re-activating the mic each time.
- Be specific with service names - speak them clearly. Tessa uses your Grid's service catalog to match what she hears.
- Use Escape to interrupt long responses and redirect. In VR, this is mapped to the controller menu button.
- Combine voice with navigation - teleport to a service platform and ask "What's wrong here?" Tessa uses your position as context.
- Mute in noisy environments - use
/muteto switch to text-only responses if background noise is an issue.
Combining Voice and Text¶
Voice and text work interchangeably in the same conversation:
- Speak a question, then type a follow-up
- Type a question, then speak "Tell me more"
- Switch between modes freely - conversation context is preserved regardless of input method
Chat history captures both voice and typed interactions, and everything is saved across sessions.
Related¶
- Tessa Overview - All capabilities and commands
- Prompt Guide - Example prompts (work with both voice and text)
- Multimodal Vision - Combine voice questions with pasted images