General Questions
What is CARTER?
What is CARTER?
CARTER is an experimental voice AI built with Cartesia’s Sonic model and deployed on Solana. It demonstrates real-world voice AI integration with genuine emotional expression and an unfiltered, creative personality.
Is CARTER production-ready?
Is CARTER production-ready?
No. CARTER is intentionally experimental and unstable. It’s built as a demonstration of Cartesia’s capabilities and to inspire developers, not for production use as-is.
Can I use CARTER in my project?
Can I use CARTER in my project?
CARTER itself isn’t meant for direct integration, but you can use the same technology (Cartesia’s Sonic model) to build your own voice AI. Check our Developer Guide.
Why does CARTER have an unfiltered personality?
Why does CARTER have an unfiltered personality?
To showcase the creative potential of AI when guardrails are removed. It demonstrates that voice AI can have genuine personality beyond standard chatbot responses.
Technical Questions
What technology powers CARTER's voice?
What technology powers CARTER's voice?
CARTER uses Cartesia’s Sonic model, which provides:
- Sub-6ms latency
- Real emotional expression
- Multiple voice options
- Streaming audio generation
- Natural-sounding speech
Why Solana blockchain?
Why Solana blockchain?
Solana provides:
- Ultra-fast transaction processing
- Scalability for high-volume requests
- Low transaction costs
- Decentralized infrastructure
How does CARTER achieve such low latency?
How does CARTER achieve such low latency?
Through a combination of:
- Cartesia’s optimized Sonic model (6ms average)
- WebSocket connections for streaming
- Solana’s fast blockchain
- Efficient audio encoding (PCM)
Can I access CARTER's source code?
Can I access CARTER's source code?
The integration patterns and architecture are documented in our Developer Section. For specific implementation details, contact the team.
Usage Questions
Why do I get rate limit errors?
Why do I get rate limit errors?
Rate limits are intentional to demonstrate real-world API constraints. They show how to handle limits in production applications.If you encounter limits, try:
- Waiting a moment and retrying
- Using the Cartesia playground: https://play.cartesia.ai/text-to-speech
How do I use voice input?
How do I use voice input?
- Click the “VOICE INPUT” button
- Allow microphone access when prompted
- Speak clearly into your microphone
- Wait for CARTER’s voice response
Why isn't CARTER responding?
Why isn't CARTER responding?
Common reasons:
- Rate limits exceeded (wait and retry)
- Server overload (try later)
- Network connectivity issues
- Browser compatibility (use Chrome/Firefox/Safari/Edge)
Does CARTER remember previous conversations?
Does CARTER remember previous conversations?
Currently, CARTER maintains context within a single session but doesn’t persist conversations across sessions. This is intentional for the demo.
Developer Questions
How can I build something like CARTER?
How can I build something like CARTER?
Follow these steps:
- Sign up for Cartesia API access
- Review our Integration Guide
- Study the Code Examples
- Check the Voice API Reference
- Start with simple TTS, then add streaming
What programming languages are supported?
What programming languages are supported?
Cartesia provides SDKs for:
- JavaScript/TypeScript (
@cartesia/cartesia-js) - Python (
cartesia)
How much does Cartesia cost?
How much does Cartesia cost?
Cartesia offers multiple pricing tiers:
- Free: Limited requests for testing
- Pro: Higher limits for production
- Enterprise: Custom limits and support
Can I clone CARTER's voice?
Can I clone CARTER's voice?
CARTER uses a pre-built Cartesia voice. You can clone your own voices using Cartesia’s voice cloning feature. See our Voice API Reference for details.
What's the difference between SSE and WebSocket?
What's the difference between SSE and WebSocket?
Server-Sent Events (SSE):
- One-way server→client streaming
- Simpler to implement
- Good for web applications
- Two-way communication
- Lower latency
- Better for real-time chat
Troubleshooting
Microphone permission denied
Microphone permission denied
Solution:
- Check browser settings (Settings → Privacy → Microphone)
- Ensure HTTPS connection (required for mic access)
- Try a different browser
- Check system microphone permissions
No audio output
No audio output
Solution:
- Check system volume
- Verify browser audio isn’t muted
- Try headphones/speakers
- Check audio output device in system settings
High latency responses
High latency responses
Possible causes:
- Network connectivity
- Server load
- Using MP3 instead of PCM format
- Not using WebSocket for real-time
API errors
API errors
Common errors:
- 401 Unauthorized: Invalid API key
- 429 Rate Limit: Too many requests
- 500 Server Error: Temporary issue, retry
Project Questions
Who built CARTER?
Who built CARTER?
CARTER was built as a demonstration of Cartesia’s Sonic model capabilities, showcasing real-world integration patterns for voice AI.
Is CARTER open source?
Is CARTER open source?
The architecture and integration patterns are documented and shared. Check the Developer Section for implementation details.
How can I contribute?
How can I contribute?
Currently, CARTER is a reference implementation. To contribute to the ecosystem:
- Build your own voice AI projects with Cartesia
- Share your implementations
- Provide feedback on the documentation
Will CARTER be updated?
Will CARTER be updated?
CARTER serves as a snapshot of current Cartesia capabilities. Updates may occur to showcase new features or improvements.
Getting Help
Developer Docs
Technical documentation
Cartesia Docs
Official Cartesia documentation
Try CARTER
Experience the interface
Contact
Reach out on X (Twitter)
Still have questions? Reach out on X (Twitter) or check the Cartesia documentation.