New Kitten TTS models released

KittenML has launched three new open-source text-to-speech models, ranging from 14 million to 80 million parameters. These models are designed for on-device applications, emphasizing small size and expressivity. The smallest 14M variant achieves state-of-the-art performance for its size, showcasing significant advancements in efficient AI for speech synthesis. This release builds upon previous iterations, offering developers more options for integrating high-quality, compact TTS capabilities into their projects, potentially enabling more sophisticated voice interactions on a wider range of devices.