Cartesia Sonic-3

Sonic-3.6 is Cartesia's real-time text-to-speech API designed for AI agents and interactive applications. It emphasizes low latency, natural pacing, and expressive voices, including laughter and nuanced emotion. The model targets production environments where speed and voice quality matter for customer interactions, agents, and multimodal experiences.

The Sonic-3.6 model supports streaming playback with very low latency, enabling conversations to feel natural as audio is produced. It is positioned for enterprise and regulated deployments where speed, reliability, and security are priorities.

Cartesia positions Sonic-3.6 as a production-ready TTS engine that can be deployed in various environments, including on-premises and in VPCs, with a focus on latency, reliability, and language coverage.

The pricing and deployment options indicate a multi-tier offering, including free and paid plans, with enterprise settings available. Cartesia also references compliance and data-residency options, suggesting suitability for regulated industries and sensitive data contexts.

Use cases

• Voice-enabled AI agents for marketing, sales, and support
• Customer support and self-service
• Multilingual enterprise communications