May 2025 Summaries

Cartesia is enhancing its voice AI platform for developers by introducing new features that improve collaboration and visibility. The Organizations feature allows teams to work more efficiently by providing shared access to API keys, custom voices, and centralized billing under a single account, with unlimited team member additions available. Meanwhile, Dashboards offer real-time insights into usage and performance, allowing users to monitor credit balances, view concurrency of voice generations, and check WebSocket connections to better manage resources. Additionally, Cartesia introduces Ink, a speech-to-text model designed for real-time, multimodal intelligence across various devices.

Starting today, users can create Professional Voice Clones (PVCs) using Sonic, a highly efficient text-to-speech model, which allows for precise replication of tone, cadence, style, and environment in voice data. Previously costly and inaccessible, PVCs are now more affordable and scalable, available on the Startup plan and above, thanks to new model infrastructure that reduces costs without compromising quality or latency. This service is ideal for businesses creating virtual avatars, AI agents, or private voice libraries, with a flexible credit system enabling unlimited PVCs. The Startup plan offers 1.25 million credits per month, sufficient for creating up to 15 voices annually, with training and generating speech incurring specific credit costs. PVCs are especially beneficial for industries such as healthcare and hospitality, allowing them to develop distinct, branded voices, and platforms for personal AI avatars can create lifelike voice clones. Available in 15 languages, PVCs can be accessed through the company's Playground UI and API, with more information provided in their documentation.