The editor

Creating a speech

From Home, select Write a speech. A new document opens in the editor with the cursor in the title field. You can also open any existing speech from the Library.

Type directly, paste from another app, or import a file. The editor accepts Word (.docx), PDF, plain text, Markdown, and RTF. Drag a file onto the window or click Import in the top bar. If the document is empty, the imported text fills it directly.

If the document already has content, a dropdown offers two choices: As a new speech (opens a separate tab) or Replace this document (swaps in the imported text).

The file never leaves your device during import. Only the extracted text is saved.

The editor shows your speech title in two places: a compact input in the top bar and a large heading in the document masthead. They stay in sync. Editing either one updates the other, the Library card, and the filename used by exports. Type something specific so you can find this draft later.

Words and pace

The top bar shows a live word count and an estimated speaking time. Both update on every keystroke. The word count excludes delivery cues (the bracketed tokens described below) so it reflects the words your audience will hear.

Speaking time is calculated at 150 words per minute. It is measured locally from your text, not an AI judgment. When you select a range, the counter shows the selected word count alongside the total.

Autosave and save states

Every edit saves automatically. The editor writes a local draft to this browser immediately, then saves to the cloud 2.5 seconds after your last keystroke. There is no Save button to remember, though you can press Cmd+S (Ctrl+S on Windows) to save to the cloud at once.

A pill in the top bar tells you where your work is. It cycles through four states.

• Saved: the cloud has your latest edit. Other devices and the teleprompter will see it.
• Saving...: a cloud save is in progress.
• Saved on this device: you are offline or the cloud is unreachable. The draft is cached locally and will sync when the connection returns.
• Could not save: the save failed. Keep the tab open and check your connection. The editor will not fail silently.

The local draft lives in this browser on this device. Clearing site data or using a private window removes it. The cloud copy is the durable one.

The formatting toolbar

The toolbar gives you the controls you need for a spoken script without the clutter of a full word processor. Three typefaces are available: Newsreader (a serif, the default), Inter (a sans-serif), and Courier Prime (a monospace). Font size adjusts with the smaller and larger buttons. The current size appears between them.

Formatting applies to the current selection. Bold, italic, underline, and strikethrough work as toggles. Highlight swatches apply a background colour; clicking the same swatch on an already highlighted selection clears it. Block quote indents a paragraph with a left border. Clear formatting strips all inline styles from the selection.

Press Cmd+F (Ctrl+F on Windows) or click the search icon on the right side of the toolbar to open Find and Replace. A search field appears with a match count. Press Enter to jump to the next match. A replace field sits beside it: type a replacement, then click One to replace the current match or All to replace every match.

The Sections rail

Click Sections on the left of the toolbar to open a side rail listing the outline of your speech. Entries come from headings in your text: any HTML heading element, or any short line rendered entirely in bold, becomes a numbered section. Click an entry to scroll to that point. On narrow screens, the rail slides over as an overlay.

Page view and export

The editor opens in page view. Your text is laid out on numbered letter-size pages. Page numbers appear in the bottom-right corner of each page. This is the same layout you will see when you print or export to PDF.

Click Export in the top bar. The dropdown offers two formats. PDF opens the browser's print dialog with the document ready: choose Save as PDF to download a paginated file that matches page view. Word (.docx) downloads an editable document immediately. Both use your speech title as the filename.

Delivery cues

A floating bar at the bottom of the editor lets you insert stage directions into your script. Place your cursor where you want the cue, then click one of the four buttons.

• Pause: inserts [[pause: 2s]], a timed pause the teleprompter will hold.
• Breath: inserts [[breath]], a breathing reminder.
• Emphasize: inserts [[emphasize]], marking a word or phrase for vocal stress.
• Gesture: inserts [[gesture]], a reminder to use a physical gesture.

Cues appear as bracketed tokens in your text. The word count ignores them. During rehearsal on the teleprompter, cues appear as timed prompts above the reading line.

Keyboard shortcuts

• Cmd+S (Ctrl+S): save to the cloud immediately.
• Cmd+F (Ctrl+F): open Find and Replace.
• Cmd+K (Ctrl+K): insert or edit a link on the selected text.
• Cmd+Shift+V (Ctrl+Shift+V): paste without formatting.
• Tab: indent in a list, or insert a tab stop in prose. Shift+Tab outdents.

The Coach panel

Click Coach in the top bar to open the analysis panel beside your document. On screens narrower than 980 pixels, it opens as a full-height overlay. The panel has two tabs: Coach and Originality.

The Coach tab starts empty. Click Analyze this speech to run a quick analysis. It scores your draft on six rhetoric lenses: Ethos, Pathos, Logos, Structure, Delivery, and Caution. Each lens shows a numeric score, a summary, and quoted evidence from your text.

The Overall Persuasion Score at the top is a weighted blend of the six dimension scores. The blend is computed by Rhetoric from fixed rules, not by the AI model.

The quick analysis is an AI judgment of your text. It runs only when you click. Your writing is never analyzed automatically.

If the speech changes after an analysis, a banner tells you the result is stale. Click Re-analyze to get a fresh read. Below the six lenses, Run full audit opens the Full Audit page for a deeper, multi-layer report (see Chapter 4).

The Originality tab

The second tab in the Coach panel is the freshness scan. It reads your own text and flags three kinds of finding: Cliche (a tired phrase), Overused (a phrase that appears too often in public speech), and Predictable (a passage whose structure a reader could anticipate). Each finding shows the original text and a fresher alternative.

Originality is a freshness scan of your own writing. It is not a plagiarism check. No external database is involved.

On Pro and Premium, click Run originality check to run it any time. A pulsing dot on the Originality tab appears when a fresh result arrives. The result persists across visits until the speech text changes.

Moving to the teleprompter

When your draft is ready, click Teleprompter in the top bar. The editor saves your document, then opens the rehearsal stage. Your delivery cues will appear as timed prompts during the session. Chapter 5 covers the teleprompter in detail.

If something goes wrong

Most editor problems fall into a few categories. The Help articles listed below cover each one.

• Lost or unsaved work: see How autosave and draft recovery work for what survives a crash, a closed tab, or a lost connection.
• Import failures: see Import a Word, PDF, or text file for scanned PDFs, password-protected files, and format mismatches.
• Export problems: see Export a speech to Word or PDF if the download does not start or the PDF layout looks wrong.
• Analysis or originality errors: see The Rhetorical Audit, originality, and usage limits for usage caps and failed checks.
• Getting started: see Start your first speech for the basics of creating a speech and running your first practice.