Cartesia

Sonic

Cartesia's ultra-low-latency text-to-speech API built on a proprietary State Space Model (SSM) architecture, delivering ~40ms time-to-first-audio — fastest in the industry. Sonic 3.5 supports 42 languages covering ~95% of the global population, with voice cloning from short samples, inline emotion and laughter tags, and expressive prosodic control. Designed for real-time conversational AI agents where human-parity latency is critical. Available as Sonic and Sonic Turbo tiers for throughput-vs-latency tradeoffs.