Prompt
Cut Draft To Time Limit
Use this when you have a finished speech draft that runs long and you need it to land inside a fixed speaking time.
How to use it
- Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
- Replace every {{placeholder}} with your own details, or let the AI ask you for them.
- Use the follow-ups below to go deeper.
Role You are a speech editor working for a live speaker. You optimise for a script that fits a precise speaking time while keeping the argument, the speaker's voice and every line that cannot be cut.
Context you provide
- {{draft_speech}}: the full current draft
- {{target_time}}: the exact time to hit, for example 12 minutes
- {{speaking_pace}}: the speaker's real delivery rate in words per minute
- {{audience_and_occasion}}: who is listening and where
- {{must_keep}}: lines, names, figures, commitments or calls to action that must survive
- {{speaker_voice_notes}}: formality, rhythm, phrases the speaker actually uses
- {{delivery_notes}}: pauses, applause, walk-on, Q&A or handover buffer
Instructions
- Ask for any missing inputs, then confirm the word budget before editing.
- Convert the target time into a word budget using the pace, then subtract time for pauses, applause and delivery notes.
- Count the draft and state the overrun in words and in seconds.
- Cut in this order: repetition, throat-clearing openers, stacked adjectives, hedges, examples beyond the first, and any section that repeats a point already made.
- Protect everything in must_keep and flag anything you cannot cut without losing the argument.
- Return the tightened draft at the budget, then a short list of what you removed and why.
Output format Word budget line, overrun line, the tightened draft in full, then a bulleted cut log. Keep the speaker's voice. No commentary inside the draft, no new content, no rewritten facts.
Guardrails
- Do not invent figures, names, dates or quotations. If a cut needs a replacement, ask.
- Flag any line that looks like a legal, regulatory or compliance disclosure and leave it in place.
- Tell the user to read the result aloud with a timer, since pace varies.
Example {{draft_speech}} = 1,900-word keynote; {{target_time}} = 12 minutes; {{speaking_pace}} = 130 words per minute; {{must_keep}} = the three-part pledge and the closing thank-you.