How to write a script by talking instead of typing (voice dictation)

Typing out a script is the slowest part of making a voiceover for a lot of people — talking it out loud is faster and more natural. Dictate turns your speech into editable text using your browser's built-in speech recognition, free, with no upload or account needed.

How it works

Click Dictate, hit the microphone, and talk — your words appear as live text while you speak. Stop whenever you're done, then choose Insert (at your cursor), Append (at the end), or Replace (the whole script) to commit it. Nothing is added to your script until you choose one of those — the dictated text sits in an editable box first.

Review before you commit — this matters

Speech recognition is convenient, not perfect. Before inserting dictated text, check it for:

• Spelling — names, brand terms, and less common words are the most likely to come out wrong. Your browser underlines misspelled words automatically in the dictation box, the same as any text field.
• Punctuation — dictation is best-effort here. It won't reliably add commas, periods, or paragraph breaks the way you'd type them; run through the text and add what's missing before committing.
• Meaning — homophones ("their"/"there," "to"/"too") and mis-heard words don't always look wrong, only sound wrong when the voiceover plays it back. A quick read-through catches these before they reach the final audio.

Why this is a deliberate design, not a shortcut

The dictation box is intentionally a holding area, not a direct-to-script pipe — you always see and can edit the raw transcript before it becomes part of your actual project. That pause is what makes dictation safe to use for a real script instead of just a rough note.

Browser support

Dictate uses the Web Speech API, which currently works in Chrome and Edge. Where it isn't supported, the button doesn't appear — typing your script works everywhere regardless.