Elmren Voice

Tonight the usual reader is away on a work trip, with the bedtime story stopped at chapter nine. Elmren Voice turns texts and books you already have into spoken audio. Each word lights up as the voice reads it, so a child can follow along. The speech is generated on the Mac itself, and the privacy page says text, recordings and finished files stay there.

Text can be pasted in or imported from EPUB, PDF, Word and subtitle files, and scanned PDFs need a text layer. Most of the 64 built-in voices come from the open Kokoro and Qwen models. Cloning a voice takes a recording of six to ten seconds, which the app checks for noise and transcript errors. The terms allow cloning only of voices you own or have explicit permission to use.

Finished work leaves the app as chapter audio, a chaptered M4B audiobook, read-along HTML pages or an MP4 video. The first launch downloads the engine and voice packs, from about 700 MB up to a few gigabytes. After that, daily work runs offline on a Mac with Apple silicon and macOS 13 or later.