Drop in a sermon transcript, a manuscript, an interview, or a YouTube script. Voices strips the timestamps, finds every speaker, casts a different voice for each one, applies a delivery style with real conviction behind it, adds a soundscape that carries the weight of the moment, and turns it into a finished audio track ready to go forth.
Pick a mode above first โ Audiobook narrates everything, YouTube Script auto-detects scene directions and [MUSIC]/(SFX) cues and turns them into automatic stings and soundscape switches instead of reading them aloud.
SRT/VTT cues, "Name:" labels, and interview formats are detected. Every speaker gets their own voice, pitch, and pace.
A revival-preacher delivery style sits right alongside Movie Trailer, Documentary, News Anchor, True Crime, Hype, Horror, and Comedic โ real conviction in the pacing and the pauses.
YouTube Script mode reads scene directions and [MUSIC]/(SFX) cues and auto-triggers whooshes, risers, impacts, and soundscape switches โ nothing gets read aloud by mistake.
15 live-synthesized beds โ cinematic, epic orchestral, revival organ, corporate, lo-fi, news desk, rain, ocean, fireside, and more โ auto-ducked under the voice.
Amplified, Cathedral, Radio, Telephone. Premium neural voices supported with your own ElevenLabs key.
Bring your own Claude key for AI-assisted chapter polish, titles, show notes, and caption ideas โ always in the same on-fire voice.
Double-click any line right in the script to fix it on the spot. Smooth Narration mode softens the pacing for a more natural read.
Record the full mix, render true MP3/WAV with premium voices, or export TXT/PDF/EPUB manuscripts and SRT captions.
Share straight to your phone's apps, post to X/Facebook, or install Voices itself as an app on your home screen.
๐ก Double-click any line in the script to edit it instantly, right where it sits.
In YouTube Script mode, [MUSIC]/(SFX) cues in your script trigger these automatically during playback and recording โ click any one to preview it now.
Uploaded tracks act as a background bed just like the synthesized soundscapes above โ pick one to loop under your narration. They live in this browser tab only and aren't included when you Save project.
Effects apply to the soundscape mix and to premium-voice renders. Built-in device voices play clean by design of the browser.
One click. Your browser will ask to share this tab with audio โ allow it, and Voices performs the chapter and captures voice + soundscape into a downloadable audio file. Works best in Chrome or Edge on desktop.
Also want it as a different file type?
Plug in your own ElevenLabs API key to render true studio-grade MP3 chapters with ultra-realistic neural voices, mixed with your soundscape. Your key stays in this browser only.
Timing is estimated from reading pace and your delivery style โ built for YouTube captions, fine-tune in your editor if needed.
Bring your own Anthropic API key (console.anthropic.com โ API Keys) to unlock AI-assisted polish, titles, show notes, and captions โ powered by Claude, in the same on-fire voice as the rest of this studio. Your key stays in this browser only; use this on a private deployment, since a browser-side key is visible to anyone with dev tools open.
Polish opens the result in the chapter editor for your review โ nothing is applied until you click Save changes there.
Shares your last recorded or rendered audio straight to your phone's share sheet โ Messages, WhatsApp, Instagram, wherever it needs to go. Record or render a chapter first.
X and Facebook open a prefilled post in a new tab โ since browsers can't attach a local file automatically, download your audio first and attach it there. Connect AI Studio above for tailored ๐ฅ caption ideas instead of the default one.
Every chapter, in one document. Use # Chapter Title on its own line to mark chapter breaks โ everything else re-detects speakers (and cues, in YouTube Script mode) exactly like a fresh import.