Multi-Engine Text to Speech provides a comprehensive voice synthesis workflow for generating natural audio from plain text or formatted scripts. It defaults to high-definition MiniMax TTS with support for custom voice cloning, word-level subtitle timestamp generation, and context-aware delivery profiles, while also offering offline Kokoro TTS (102 voices via local Docker) and Microsoft Edge TTS as versatile fallbacks. Ideal for video narrations, tutorial voiceovers, podcasts, and automated multimedia production, the tool automatically removes script cues, BGM markers, and timestamps before synthesizing clean, natural bilingual Mandarin and English speech.
주요 기능
01MiniMax voice cloning support with integrated cost pre-checks, sample validation, and secure local profile management
02Automatic script sanitization to strip stage directions, BGM markers, timestamps, and markdown formatting
03Word-level subtitle timestamp export into structured JSON for automated video synchronization
04615 GitHub stars
05Multi-engine support for high-definition MiniMax TTS, offline Kokoro TTS v1.1-zh, and Edge TTS
06Delivery consistency profiles such as friendly tutorial and commercial narration for unified video pacing
사용 사례
01Generating natural, consistent voiceovers for video tutorials, social media shorts, and corporate explainers
02Converting structured podcast and video scripts into spoken audio with stage notes automatically removed
03Synthesizing speech locally and offline without external API latency using Docker-hosted Kokoro TTS