Podcast Intro Voice: Make One With AI in 5 Steps

A podcast intro voice is the short spoken line that opens an episode and names the show. Keep the spoken part to about 15 to 30 seconds, which is a guideline from one chart analysis rather than a rule. AI can voice the script in minutes, but you still write the line, mix it to Apple's loudness target and meet the platforms' disclosure rules.

How long should a podcast intro be?

The Podcast Host analysed the top 20 shows on the US Apple Podcasts charts. The article was last updated January 29, 2025, and it does not state the exact snapshot date. Its findings:

7 of the 20 shows had intros longer than 30 seconds.
6 of the 20 had only a jingle of about 3 seconds, plus a "Welcome to" line at most.
Every show had a distinct jingle.
The longest shows, at 2 to 3 hours, started their content within a second or two.
A host-chat intro of 5 to 10 minutes drew negative reviews.

The author calls 30 seconds a "sweet spot" borrowed from advertising, and advises checking listener drop-off in Apple Podcasts Connect. Read the full analysis at The Podcast Host's intro formula.

To turn seconds into words, use the audiobook pace of about 155 words per minute from ACX's reading rate. That gives roughly 39 words for 15 seconds and 78 words for 30 seconds. Treat these as estimates and adjust for your own delivery.

Podcast intro script template

A good intro usually has four parts. The seconds below are a suggested budget built from the chart analysis, not a figure from a source.

| Element | Purpose | Suggested seconds | Example | |---|---|---|---| | Jingle | Signals the show | About 3 | Your theme, with no voice | | Welcome line | Names the show | 3 to 5 | "Welcome to [Show Name]." | | Promise | Says what this episode covers | 5 to 10 | "This week we look at ..." | | Optional sponsor | Funds the show | Keep it short | One line, read clearly |

The same analysis notes that calls to action work best mid-episode or near the end, not at the top.

How to make a podcast intro voice with AI in 5 steps

Write a 15 to 30 second script. Use the word range above, and keep the welcome line in the first sentence.
Choose a voice. OndaLabs offers 30 prebuilt voices usable in every supported language. You can also clone a voice from a recording you have permission to use, or design one from a text description for 1,500 credits.
Set the delivery. Describe the tone in plain words, and add supported expressive tags where they help. Onda Flash is the most expressive model and is available on paid plans.
Generate and listen. Create a few takes and listen on phone speakers as well as headphones. Keep the one that sounds like the show.
Add music and mix to the target loudness. Place the voice over your jingle in an editor, then measure and export to the specifications below.

The OndaLabs text to speech tool is the place to run steps 2 to 4. For the jingle, music generation is available on Creator and up, at 2,500 credits per clip or 5,000 credits per song.

Add music and mix to the right loudness

Apple's RSS guidance and Spotify's creator guidance differ, so check each separately. The table shows the figures we could verify.

| Specification | Apple Podcasts (RSS) | Spotify (creator page) | |---|---|---| | Target loudness | About -16 LKFS, plus or minus 1 dB | Not stated on the page | | True peak | No higher than -1 dB FS | Not stated on the page | | Accepted formats | MP3 or AAC (AAC in an MP4 container) | MP3, M4A or WAV | | Bit rate at 44.1 or 48 kHz | Mono 64 to 128 kbps; stereo 128 to 256 kbps | Not stated on the page | | Channels | Mono or stereo | Mono or stereo |

Apple specifies loudness with ITU-R BS.1770-5 and says to apply loudness processing before encoding (Apple Podcasts audio requirements). Apple Podcasts Connect uploads use a separate set of figures. They accept WAV, FLAC or MP3 only, and recommend mono MP3 at 96 to 128 kbps.

Mono files need a special note. A mono file measures about 3 LU quieter than the same audio in stereo, so it needs roughly -19 LUFS to sound as loud as -16 LUFS stereo. Some analysers do not make this adjustment (Podnews: LUFS and LKFS for podcasters). Spotify and YouTube are often cited at -14 LUFS, but Spotify's own page does not state that figure, so treat it as a common reference rather than a requirement.

OndaLabs does not describe loudness processing, so measure and adjust the levels in your audio editor.

AI disclosure and voice rights

Disclosure and rights are where AI intros cause the most trouble. Three points matter:

Apple requires disclosure. Guideline 1.11 requires creators to disclose AI-generated voices, hosts and replicas prominently, in the content and in the metadata for each episode and show. A spoken line plus a metadata note is the safest reading. That reading is our interpretation of the guideline, so check the current text.
Spotify bans impersonation. On May 19, 2026, Spotify said it will remove shows and content that impersonate another creator or host's likeness without permission, including through AI voice cloning (Music Business Worldwide).
Generated audio may not be protected. The US Copyright Office says a purely AI-generated element may not be protected, while a human-written script and your own arrangement can be (US Copyright Office, Part 2 report). If your intro music matters to your brand, do not assume you can stop others from reusing a generated clip.

Clone only your own voice, or a voice you have permission to use. Cloning another person's voice without permission is the kind of impersonation Spotify removes.

Common mistakes

A long host chat before the content. The chart analysis links long intros to negative reviews.
Skipping the metadata note. Apple expects disclosure in the metadata as well as in the audio.
Mixing mono and stereo without checking. Mono can sound quieter than expected at the same target.
Assuming a generated jingle is yours to protect. Generated material may not be protected.
Cloning a voice without permission. This is the fastest route to a removed show.

What the intro costs on OndaLabs

The free plan is a low-risk way to test an intro. It includes 2,000 credits a month on Onda Lite, which is about 2 minutes of audio. Each character costs one credit on Lite, so count the characters in your script to estimate the cost before you generate. Creator includes 3 custom voices, and Pro includes 10. OndaLabs pricing lists the current plan details.

Speech to text costs 50 credits per minute, so you can transcribe the finished episode for show notes. Commercial use is included from the Starter plan at $9 per month, which matters if the podcast is sold or sponsored.

Frequently asked questions

How long should a podcast intro be?

Many successful shows keep the intro short. In one analysis of the top 20 US shows, 6 of 20 used only a jingle plus a welcome line, while 7 ran longer than 30 seconds. A 15 to 30 second spoken intro is a practical range, but test it with your own listeners and check drop-off in Apple Podcasts Connect.

What should I say in a podcast intro?

Say the show name early, then one sentence about what the episode covers. Keep sponsor reads brief, and move calls to action to the middle or end of the episode. A short, clear line works better than a long greeting.

Can I use an AI voice for my podcast intro?

Yes, as long as you follow the platform rules. Apple requires disclosure of AI-generated voices in the content and in the metadata for each episode and show. Spotify removes impersonation without permission, so use a voice you have rights to.

Do I have to disclose AI voice on Apple Podcasts?

Yes. Apple's guideline 1.11 requires prominent disclosure of AI-generated voices, AI hosts and AI replicas, in the content and in the metadata for each episode and show. The guideline is the requirement, so read it before you publish.

What loudness should my podcast intro be (LUFS)?

For Apple Podcasts, the RSS guidance targets about -16 LKFS with a tolerance of plus or minus 1 dB, and a true peak no higher than -1 dB FS. A mono file needs about -19 LUFS to sound as loud as -16 LUFS stereo. Measure the final file in your editor before you upload.

Can I clone my own voice for the intro?

Yes, if you record a sample you have permission to use. Cloning is useful for a consistent host voice across many episodes. Confirm the consent and rights for every voice you use, and keep a record of them.