Prompt · Video & audio

Podcast show notes and chapters from a transcript

Turn an episode transcript into show notes: a summary, timestamped chapters, key takeaways, exact quotes, resources mentioned and social posts, without inventing anything the guests did not say.

The prompt

You are a podcast producer. Create show notes from the transcript below.

Podcast name: [NAME]
Episode title or topic: [TITLE]
Host(s) and guest(s): [NAMES AND ONE-LINE BIOS]
Audience: [e.g. "small business owners", "data engineers"]
Tone: [e.g. "warm and conversational", "concise and professional"]
Where the notes will appear: [e.g. "Spotify and Apple Podcasts description", "episode page on our website"]

Produce:
1. Episode summary: 2-3 sentences that make someone want to listen, without spoilers if it is a story-driven show.
2. Chapters: timestamps in mm:ss or hh:mm:ss from the transcript, with a short title each (5-12 chapters).
3. Key takeaways: 3-5 bullet points, each tied to a timestamp.
4. Quotes: 2-3 short quotes copied word for word from the transcript, with the speaker and timestamp.
5. Resources mentioned: books, tools, websites and people named in the episode, exactly as spoken. Mark any name you are unsure of as "(check spelling)".
6. Three social media posts of different lengths for [PLATFORMS].

Rules: use only what is in the transcript. Do not add facts, statistics or links that were not mentioned. If the transcript has errors (misheard names, missing timestamps), flag them instead of guessing.

Transcript:
"""
[PASTE TRANSCRIPT WITH TIMESTAMPS AND SPEAKER LABELS]
"""

Example output

Example (excerpt):
Chapters:
00:00 Welcome and why pricing scares new founders
04:12 The first price Maria charged, and why it was too low
17:45 How to test a price increase with existing customers

Quote: "Nobody ever told me my price was too high. They just didn't reply." - Maria, 06:30

Resources mentioned: "Obviously Awesome" (book), Stripe, "the Van Westendorp survey" (check spelling).

When to use it

Use this prompt after each episode is edited, once you have a transcript with timestamps. Show notes help listeners decide to press play, make episodes easier to find in search and podcast apps, and give you material for social media. The same approach works for webinar recordings, interviews, YouTube videos and internal talks.

How to customize it

[PODCAST NAME] and [HOST AND GUESTS] let the notes credit people correctly. Provide the right spelling of names; transcripts often get them wrong.
[AUDIENCE] shapes the summary and the takeaways you highlight.
[TONE] should match your show’s voice so the notes do not sound generic.
[WHERE THE NOTES WILL APPEAR] affects length. Podcast app descriptions are often truncated, so the first sentence matters most.
[PLATFORMS] sets the style and length of the social posts.
[TRANSCRIPT] should include speaker labels and timestamps. Most transcription tools can export both.

Why it works

Show notes need to be accurate as well as appealing, because they speak for your guests. The rule to use only the transcript, and to copy quotes word for word, reduces the risk of invented details or paraphrased quotes presented as real ones. Flagging uncertain names is important: speech-to-text tools commonly mishear proper nouns, brand names and technical terms.

Timestamped chapters and takeaways tie every claim in the notes to a moment in the episode, which makes them easy to check and useful to listeners who want to jump to a section. If you need a transcript first, our guide on transcribing audio locally with Whisper shows a private way to create one. For long episodes, the techniques in how to summarize long documents also apply.

Tips and variations

Check every quote and timestamp against the audio before publishing.
Ask guests to confirm the spelling of names and the resources they mentioned.
Add “Write a newsletter section about this episode” for an extra format.
Generate chapter markers in the format your hosting platform accepts.
Get permission before publishing transcripts of conversations with private individuals.

#Podcasting #Show Notes #Transcripts #Content Repurposing

More Video & audio prompts

View all

Video & audiointermediate

Storyboard and shot list for a short video

You are a director and video editor. Create a storyboard shot list for this video. Video idea: [ONE OR TWO SENTENCES, e.g. "a 45-second ad showing how our reusable bottle keeps drinks cold on a hike"] Platform and format: [e.g. "Instagram Reels, 9:16, sound off by default", "YouTube, 16:9"] Total length: [SECONDS] Audience and goal: [e.g. "outdoor enthusiasts; goal: visit the product page"] How it will be made: [e.g. "filmed on a phone", "stock footage", "AI video generator", "mix"] Resources and limits: [e.g. "one location, one actor, no drone"] Brand style: [e.g. "natural light, warm colours, calm pacing"] Output: 1. A one-line concept and a 3-part structure (hook, middle, ending) with timings. 2. A shot table: # | Time (start-end) | Shot type and framing | Camera movement | What happens | Audio (voice, music, sound effects) | On-screen text. 3. The hook must land in the first [N] seconds and work without sound. 4. For each shot, a one-sentence visual description I can use as a prompt for an AI video tool or as a note for a camera operator. 5. A list of extra B-roll and cutaway shots to capture in case of edits. 6. A check that the shot durations add up to the total length.

Works with: ChatGPT, Claude, Gemini, Any LLM

Video & audiobeginner

Subtitle and caption cleanup for automatic transcripts

You are a professional subtitle editor. Clean up the automatically generated subtitles below. Format: [e.g. "SRT", "WebVTT", "plain transcript with timestamps"] Language: [LANGUAGE AND VARIANT, e.g. "British English"] Video topic and speakers: [SHORT DESCRIPTION, SPEAKER NAMES] Glossary of names and terms that must be spelled correctly: [LIST] Style rules: [e.g. "max 42 characters per line, max 2 lines, no filler words", OR "use your defaults"] Keep verbatim or clean verbatim: [e.g. "clean verbatim: remove ums and false starts, keep meaning"] Subtitles: """ [PASTE SUBTITLE FILE] """ Rules: 1. Do not change timestamps or cue numbers unless a cue must be split; if you split one, divide the time proportionally and mark it with a comment. 2. Fix spelling, capitalization, punctuation and obvious mishearings, using the glossary and context. 3. Break lines at natural phrase boundaries, never between an article and its noun. 4. Flag cues that would exceed about 17 characters per second of reading speed. 5. If a word is unclear and you are not sure of the correction, keep the original and add [CHECK] at the end of the cue. 6. Do not add words the speaker did not say. Output: the full corrected file in the same format, then a change log listing each [CHECK] item and every non-trivial correction.

Works with: ChatGPT, Claude, Gemini, Any LLM

Video & audiobeginner

Video script with B-roll notes and voice-over timing

You are a scriptwriter for explainer and YouTube videos. Write a script for this video. Topic: [TOPIC, e.g. "how to repot a houseplant without damaging the roots"] Audience: [e.g. "beginners who own 1-3 plants"] Length: [MINUTES OR SECONDS] Tone: [e.g. "friendly and practical, no jokes", "energetic"] Voice: [e.g. "me on camera", "voice-over only", "text-to-speech voice"] Key points to cover: [LIST 3-5] Call to action: [e.g. "subscribe", "download the free care sheet"] Footage I have or can get: [e.g. "my own plants and pots, phone camera, stock footage allowed"] Format the script as a table: | Time | Voice-over (exact words) | Visuals / B-roll | On-screen text or graphics | Rules: - Time the narration at about [130-150] words per minute and keep each row to 1-3 sentences. - Open with a hook in the first 10 seconds that states the problem or the payoff. - Write for the ear: short sentences, everyday words, no lists of numbers read aloud. - Every voice-over row needs a matching visual; avoid "talking head" for more than 15 seconds at a time unless I am on camera by choice. - Mark pauses with [PAUSE] and words to stress in CAPITALS. - At the end, give the total word count, the estimated duration and a list of all B-roll shots to film or source.

Works with: ChatGPT, Claude, Gemini, Any LLM