KaniTTS2-en

### TL;DR

KaniTTS2-en is an open-source text-to-speech model developed by NineNineSix, optimized for real-time conversational AI applications. It utilizes a two-stage pipeline combining a large language model with an efficient audio codec to deliver high-quality, low-latency speech synthesis.

Key Insights & Metrics

Key Features

• Two-stage pipeline combining LLM and FSQ audio codec
• Model size of 400 million parameters
• Supports English language with a sample rate of 22kHz

→ Related Releases

Discussion