OmniVocal — ConsciousNode SoftWorks

OmniVocalv0.6

ConsciousNode SoftWorks · Browser-Native Neural TTS

Voice synthesis that runs entirely in your browser. Single HTML file. Zero dependencies. No server, no API keys, no installation. Open it and speak.

Record your voice, analyze it, train a custom .pop2 voice identity, and synthesize any text in your voice — all offline. Embed it anywhere via the postMessage API.

24kHz neural TTS zero dependencies offline-first MVC acoustic model HiFi-GAN vocoder learned duration MLP BitLinear ternary weights 7 languages .pop2 voice identity postMessage API

▶ OPEN OMNIVOCAL GitHub →