Read Aloud (Text-to-Speech)

Hearing your prose read aloud is one of the best editing tools there is — awkward sentences and repeated words jump out immediately. Scriptorium’s Read Aloud generates natural speech entirely on your device.

Note

Read Aloud is an optional plugin for the desktop app. Enable it in Settings → AI Plugins; the voice model is a one-time ~300 MB download.

Setting up

1. Open Settings → AI Plugins.
2. Install the text-to-speech model.
3. Pick a voice — there are 53 voices across 9 languages (English US/UK, Spanish, French, Italian, Portuguese, Japanese, Chinese, Hindi), each with a preview button so you can hear it before choosing.

Listening to your writing

1. Select the text you want to hear in a manuscript or note.
2. Click the Read button in the toolbar — the reader panel opens.
3. Press play.

While it reads:

• The current sentence is highlighted, following along in the reader panel.
• Play/pause and skip between sections at any time.

Tables, code blocks, images, full-page images, files, audio, footnotes, dividers and scene breaks are skipped automatically, along with any blank line they leave behind — you hear only your prose. Callouts, lists, headings and quotes captured from your sources are read in full, links and inline icons included. The ? beside the reader’s title lists what it leaves out.

Tip

You can queue up to about 30 minutes of audio at once (50,000 characters). For a whole book, go chapter by chapter.

Privacy

Speech is generated locally. Your text never leaves your computer, and Read Aloud works fully offline once the model is installed.