Cartesia Secures $100 Million for Real-Time Voice AI, Launches Sonic-3 Model

San Francisco, CA – Cartesia, a Silicon Valley-based AI voice technology company, has announced a significant milestone, raising $100 million in funding from prominent investors including Kleiner Perkins, Index Ventures, Lightspeed Venture Partners, and NVIDIA. Alongside this substantial investment, the company has officially launched Sonic-3, its latest state-of-the-art model designed for real-time, natural conversational AI. This development positions Cartesia as a key innovator in the rapidly evolving voice AI landscape.The Sonic-3 model distinguishes itself with a reported 90-millisecond latency, support for 42 languages, and the ability to convey a full emotional range, including laughter. Unlike many traditional voice AI systems that rely on Transformer architectures, Sonic-3 is built upon State Space Models (SSMs), an approach pioneered by Cartesia founders Karan Goel and Albert Gu at Stanford’s AI Lab. This architectural choice enables the model to process conversations more efficiently, mimicking human memory by maintaining contextual awareness rather than reprocessing entire dialogue histories, resulting in faster and more natural responses.The announcement was highlighted by Cartesia CEO Karan Goel, who stated on social media, "> We've raised $100M from Kleiner Perkins, Index Ventures, Lightspeed, and NVIDIA. Today we're introducing Sonic-3 - the state-of-the-art model for realtime conversation." Goel further emphasized Sonic-3's capabilities, noting its "Breakthrough naturalness - laughter and full emotional range" and its "Lightning fast" performance. This funding and technological advancement are expected to accelerate the development and adoption of highly responsive and human-like AI voice agents across various industries.