Making a podcast intro with AI takes three steps: draft the script against a fixed word budget, voice it with a text-to-speech tool or an AI host, then save the result as one reusable asset you drop on the front of every episode. Budget about 20 minutes for the first one and roughly zero for every one after that, because a good intro is written once and reused.
The step almost every guide skips is the one that now matters most. Your intro is no longer just something people hear. Most of your audience is looking at a screen when it plays, and a music sting does nothing on a muted autoplaying feed.
TL;DR:
- A podcast intro is the fixed opening segment that tells a new listener what the show is and why this episode deserves their next few minutes.
- Length follows a word budget, not a feeling. At a normal speaking pace of 2.5 words per second, an 8-second intro is 20 words and a 30-second intro is 75.
- Build the video version, not just the audio one. Video podcast consumption passed audio-only for the first time in Q3 2025: 82% of weekly podcast consumers watch, against 78% who listen to audio only (Edison Research, The Podcast Consumer 2026).
- AI belongs in the drafting and production layers. Keep the claim about who you are human, because 62% of weekly podcast consumers already see AI as a threat to podcast credibility (same study).
- The outro has one job, and it is not "thanks for listening". It is a single instruction.
What does a podcast intro actually have to do?
An intro has one job: convert a stranger into someone who stays. It is not a table of contents and it is not a credits roll. A new listener is deciding, in a handful of seconds, whether this show is for them, and everything in the intro either helps that decision or gets cut.
That reframes what belongs in it. Your own name is not a reason to keep listening. The show's founding story is not a reason. The reason is the specific thing this show gives someone who has a specific problem or interest, said fast enough that they hear it before they leave.
Two things also get confused constantly, so separate them now:
- The branded intro is fixed. Same words, same voice, every episode. Written once.
- The cold open changes every episode. It is the hook for this particular episode, and it goes before the branded intro, not after.
Putting the branded intro first is the most common mistake in the format. You spend your best seconds on housekeeping and then ask for attention you already lost.
How long should a podcast intro be?
Short. For a video podcast clip, 5 to 10 seconds. For a full episode, 15 to 30 seconds including music. Anything past 30 seconds is a trailer, not an intro, and it should be doing more work than announcing a title.
The useful way to hold yourself to it is a word budget rather than a stopwatch, because you write words. A natural conversational delivery runs at roughly 150 words per minute, or 2.5 words per second, which is the exact rate our own generator uses to size a script before it renders. That gives you a clean conversion table:
| Intro length | Word budget | Best for |
|---|---|---|
| 5 seconds | ~12 words | A short clip on TikTok, Reels, or Shorts |
| 8 seconds | ~20 words | A video podcast clip with a title card |
| 15 seconds | ~38 words | A YouTube video podcast episode |
| 30 seconds | ~75 words | A full audio episode, music included |
| 60 seconds | ~150 words | A show trailer or season opener, not an intro |
Write to the number. If your 15-second intro draft is 60 words, it is not a 15-second intro, and speeding up the read to fit is how intros end up sounding like a disclaimer at the end of a car advert.
What should a podcast intro say?
Four things, in this order, and you can usually drop the fourth:
- Who it is for and what they get. The audience and the payoff, in one sentence.
- The show name. Once. Said clearly, because this is how people search for you later.
- The host. A first name and a reason to trust them, in a few words.
- A promise about this episode. Only if you are recording per episode rather than reusing one asset.
A 15-second template that fits the 38-word budget:
This is [show name], where [audience] get [payoff] in [format and length]. I'm [name], [one credential]. Today: [the promise].
Filled in, at 34 words:
This is The Cold Start, where solo founders get one growth tactic they can run this week in under ten minutes. I'm Dana, I've launched four products. Today: pricing pages that convert.
Notice what is not in there. No "welcome back", no "thanks for tuning in", no apology for the gap since the last episode. Those are things you say to an audience you already have, in an asset whose entire purpose is winning an audience you do not.
How do you write a podcast intro script with AI?
Give the model constraints, not a topic. The quality of an AI-written intro tracks almost entirely with how tightly you brief it, and "write me a podcast intro" produces the same beige paragraph for everyone who asks.
A brief that works:
Write 5 options for a 15-second podcast intro, maximum 38 words each. Show: [name]. Audience: [who]. Payoff: [what they get]. Host: [name, one credential]. Tone: [direct and dry, no hype]. Do not use the words "welcome", "dive into", or "join me". Lead with the payoff, not the show name.
Then run three passes, which is where the actual writing happens:
- Read it out loud. Cut every word you stumble on. Written-language sentences are the standard failure of AI drafts, and they only reveal themselves in the mouth.
- Cut to the budget. Take the best option and delete words until it hits the count. It will improve, every time.
- Check the first four words. They carry the decision. If they are "Hi everyone and welcome", start again.
The same brief-and-edit loop applies to the rest of your script, which is covered in more depth in our guide to writing a podcast script with AI.
How do you make a video podcast intro with AI?
This is the part the audio-first guides miss, and it is now the larger half of the job. Video podcast consumption overtook audio-only for the first time in Q3 2025, with 82% of weekly podcast consumers watching video against 78% listening to audio only, according to Edison Research's Podcast Consumer 2026. Platform share tells the same story from the other direction: 44% of weekly podcast consumers name YouTube as the platform they use most, ahead of Spotify at 14% and Apple at 5%, in Cumulus Media and Signal Hill Insights' Podcast Download Spring 2026 report (600 weekly podcast consumers, fielded April 2026).
An audio sting cannot survive that context. It plays under a static cover image, on a platform where the viewer is scrolling, often with sound off. So build the video version:
- Write to the shorter budget. 8 seconds, about 20 words. Video intros have to be shorter than audio ones because the viewer can leave in one thumb movement.
- Generate a host reading it. An AI host turns the script into a talking-head clip without a camera, a studio, or a second take. This is the same pipeline as producing a full clip, described in the pillar guide on how to make a podcast with AI.
- Burn in captions. Sound-off viewing is the default, so the words have to be legible without audio.
- Add the show name on screen. The one piece of branding that earns its space, because it is what people type into search later.
- Render once, reuse forever. Keep it as a single file. Prepend it to every episode. Regenerate only when the show changes.
If your show runs in more than one language, generating rather than dubbing the intro is the cheaper path, for the reasons laid out in the guide to multilingual podcasts with AI. The script is 20 words; regenerating it per language costs almost nothing, while dubbing it drags a timing problem behind it.
What should a podcast outro say?
One instruction. That is the whole brief.
Outros fail because they try to do five things: thank the audience, plug the newsletter, mention the sponsor, ask for a review, and tease the next episode. Five asks is zero asks. The person who reached the end of your episode is the most willing audience you will ever have, and splitting their willingness across five options wastes it.
Pick the single action that matters most this quarter, say it in one sentence, and say the same sentence every episode until you change the goal. Fifteen words is plenty:
If this was useful, follow the show so the next one finds you.
Repetition is the point. An outro that changes every episode teaches people nothing; one that never changes becomes the thing your audience can recite, which is exactly when they start doing it.
Where does AI not belong in your intro?
In the claim about who you are. Your audience is more comfortable with AI in production than in identity, and the gap is measurable: 69% of weekly podcast consumers approve of AI for research and brainstorming, while 62% see AI as a threat to podcast credibility, per the same Edison Research study. Megan Lazovick, a vice president at Edison Research, described that audience as "deeply engaged, fiercely loyal, and growing every year." Fiercely loyal audiences are the ones who notice.
The practical line: use AI to draft, to voice, and to render. Do not use a synthetic voice to imply a person said something they did not. If the voice in your intro is a clone of a real host, disclosure rules apply on the major platforms, and the details of what is required are in our guide to AI voice for podcasts.
FAQ
How long should a podcast intro be?
Between 5 and 30 seconds, depending on the format. Video podcast clips want 5 to 10 seconds; full audio episodes can carry 15 to 30 including music. Size it with a word budget at 2.5 words per second rather than trying to hit a time by ear.
Do I need music in my podcast intro?
For an audio show, yes, a short sting helps mark the start. For a video clip, music matters much less than captions and a face on screen, because a large share of the audience is watching with sound off. Never let the music run long enough to delay the first spoken words past about three seconds.
Can AI write a podcast intro?
Yes, and it does the drafting well when it is given constraints: a word limit, the audience, the payoff, the tone, and a list of banned phrases. What AI does badly on its own is knowing which of the five drafts is the one your audience would actually respond to. That judgment stays with you.
What is the best AI tool for making a podcast intro?
There is no single tool, because the job has three layers. An LLM like ChatGPT or Claude drafts the script, a text-to-speech tool like ElevenLabs voices it, and an AI host tool renders the video version. Tools that bundle all three save you the assembly step, which is most of the time cost.
Should the intro come before or after the hook?
After. Open with the hook for this specific episode, then run the branded intro once the listener has a reason to stay. The only exception is a show with an audience that already knows the format, where the branded intro functions as a familiar signal rather than an introduction.
Can I reuse the same podcast intro for every episode?
Yes, and you should. A fixed intro is a recognition asset; changing it resets the familiarity you are building. Rerecord only when the show name, the host, or the audience promise changes, which for most shows means once a year at most.
Write the 20 words, then see them as a finished, captioned clip with a host reading them. Make your first one for $1 and you will have a reusable intro before you finish planning the episode it belongs to.