1. Executive Summary Vocalizer 3 is a high-fidelity, concatenative and parametric text-to-speech engine designed for embedded systems, automotive infotainment, assistive technology, and enterprise applications. Released as a major iteration following Vocalizer 2, it introduced significant improvements in naturalness, voice compactness, and language support. It remains a benchmark for on-device TTS where low latency and offline operation are critical. 2. Core Architecture Vocalizer 3 combines two synthesis methods: