Podcast & Dialogue Voice Generator — Multi-Speaker Text to Speech
MWT Voice by MWT Labs lets you convert text into natural AI speech, clone voices, and professionally edit, merge, and convert audio 100% offline in your browser. All models and audio DSP tools work locally with zero server-side inference or cloud uploads required.
Audio Studio Production Tools
- Audio Editor: Waveform trim, split at playhead, peak normalize, reverse, fades, and real-time pitch & formant voice changer.
- Audio Merger: Combine and reorder multiple audio clips, configure custom silence gaps, audition clips, and export seamless tracks.
- Audio Converter: Convert between WAV and MP3 up to 320 kbps, extract lossless audio from MP4 and WebM videos.
Supported Voice Engines
- Kokoro TTS: A great choice for high-quality, natural-sounding speech, especially English (20 American & 8 British voices), with 54 expressive voices across 8 languages.
- Pocket TTS: A good choice for users who specifically want custom zero-shot AI voice cloning, supporting English, Spanish (Español), and Brazilian Portuguese (Português Brasileiro) with 8 built-in voices (~145 MB per language).
- Kitten TTS: A lightweight option (~60 MB, 8 English voices) for users who prefer a smaller model.