Cartesia is an AI voice company founded by a team of Stanford AI Lab PhDs who invented State Space Models (SSMs), the architecture underlying its products. Its flagship model, Sonic, is a low-latency speech generation model offered alongside Ink, a streaming speech-to-text model, and Line, a platform for building enterprise voice agents.
The platform supports both instant voice cloning from short audio samples and professional voice cloning through its API and web playground. Cartesia's acceptable use policy requires that users only submit their own voice recordings or those of others with explicit consent, and prohibits impersonation of any person, including celebrities and political candidates.
Models can be deployed in the cloud, on-premise, or on-device. Target use cases include customer service, healthcare, financial services, localization, and government voice applications. The company is backed by investors including Index Ventures, Kleiner Perkins, General Catalyst, and Lightspeed.
Cartesia offers a free tier with 20,000 credits per month for text-to-speech and speech-to-text. The Pro plan ($5/month) adds instant voice cloning and a commercial license, the Startup plan ($49/month) adds professional voice cloning, and the Scale plan ($299/month) provides higher volume and concurrency. Enterprise plans include custom volume pricing, DPAs, BAAs, and SSO. All plans include unlimited seats and voice slots.