What is VoxBox?
VoxBox is a private, multi-agent voice assistant. Talk or type — VoxBox listens, thinks, and responds with voice. Each agent has a distinct personality, voice, and style.
Getting Started
1. Select your profile
Choose your group (Family, Guests, Admins) and tap your name.
2. Enter your passcode
Type your 6+ character alphanumeric passcode. First-time users will be asked to create one.
3. Talk or type
Hold the mic button to speak, or type in the text box. VoxBox transcribes your speech, sends it to the AI, and reads the response aloud.
4. Attach files
Tap the paperclip icon to attach images or documents. VoxBox can analyze images, read files, and discuss their contents.
Voice Agents
AI Engines
Switch between AI providers using the engine dropdown:
Anthropic (Claude) · OpenAI (GPT) · Groq Free · xAI (Grok) · Cerebras
Each engine has different models. Use the model dropdown to pick a specific one, or leave on "Default" for the recommended model.
Available Tools
VoxBox can do more than chat. Ask it to:
Features
- Push-to-talk: Hold the mic button to record, release to send
- Text input: Type messages when voice isn't convenient
- TTS toggle: Turn voice responses on/off with the speaker icon
- Stop button: Tap the red stop button to interrupt voice playback
- File attachments: Send images and documents for analysis
- Conversation history: Your chat history persists between sessions
- GPS location: Automatic location for nearby searches and weather
- Multi-engine: Switch AI providers mid-conversation
Tips
- For the best voice experience, use headphones to avoid echo
- Allow microphone access when prompted — VoxBox needs it for voice input
- Groq is free but may be less capable than Claude or GPT for complex tasks
- Your passcode is stored securely (hashed, not plaintext)
- Guest accounts have limited tool access for safety