How Long Should an Audio Drama Episode Be?

There is no single right length for an audio drama episode, but there are working ranges: a serialized episode usually runs 10 to 25 minutes, a standalone or "feature" drama runs 30 to 60 minutes, and a kids' episode is often 5 to 12 minutes. The right number is set by three things, not by taste: the platform you publish on, the shape of your story, and how your listener actually listens.

The short answer, in ranges

Start with a target, then adjust. These ranges reflect how most independent audio drama is packaged today:

• Serialized episode (ongoing series): 10 to 25 minutes. Long enough for a full scene arc and a hook into the next episode, short enough to fit a commute or a chore.
• Standalone or feature drama: 30 to 60 minutes. A complete story in one sitting, closer to a radio play or a short film in audio.
• Kids' or bedtime episode: 5 to 12 minutes. Matched to attention span and to the routine it slots into.
• Anthology installment: 15 to 35 minutes. Each entry is self-contained, so it can run a little longer than a serialized beat without losing anyone.

To make that concrete: a typical 15-minute serialized episode is roughly 2,000 to 2,300 spoken words, or about four scenes. A 45-minute feature is closer to 6,000 to 7,000 words across eight to twelve scenes. Those are rough conversions at an ordinary narration pace; dense dialogue reads faster than descriptive narration, and heavy sound design adds real seconds that never show up in the word count. Use the word figure to sanity-check a script before you record, then trust the actual runtime once you hear it.

These are starting points, not rules. A gripping 40-minute serialized episode beats a padded 15-minute one every time. Length is a consequence of good pacing, not a substitute for it.

What actually sets an episode's length

Three practical inputs decide where your episode lands inside those ranges.

Scene count and density. Audio has to earn every location change, because the listener cannot glance at a page to reorient. A tight episode is usually three to five scenes. More than that and you are either rushing each one or asking the listener to hold too much in their head. Count your scenes before you count your minutes.

Cast size. More speaking characters means more setup: the listener needs a beat to learn each new voice. A two-hander can move fast and stay short. A scene with six characters trading lines needs room to breathe, or every voice blurs into the next. If you are unsure two voices are distinct enough to carry a long scene, that is a pacing problem worth fixing before you worry about runtime. (See what to do when two characters sound too similar.)

Listener context. Ask where the episode gets heard. A commute is 20 to 40 minutes and rewards a self-contained arc. A bedtime listen wants something short and calming that can end without a cliffhanger. A binge-friendly weekend drop can run long. Design the length around the moment, not around a spreadsheet.

Serialized versus one-shot pacing

The single biggest length decision is whether you are writing a series or a standalone, because the two want opposite things from an ending.

A serialized episode is built around a question. It opens on momentum, plays out one clear complication, and closes on a hook that makes the next episode feel required rather than optional. That structure caps naturally at 10 to 25 minutes: one strong question does not need an hour. Consistency matters more than any exact number here, so pick a target and hold it, so subscribers know what a "new episode" costs them in time.

A one-shot or feature is built around a full arc: setup, escalation, turn, and resolution, all in one sitting. That needs 30 to 60 minutes to feel complete rather than truncated. The risk is the sagging middle, so give a feature a clear act break around the one-third mark to keep the momentum honest.

If you are planning a whole series rather than a single episode, the episode length question folds into a bigger one about season arcs, episode counts, and consistency; that is covered in how to plan an audio drama season. And if your show is closer to a produced podcast than a scripted drama, the general runtime tradeoffs live in how long should a podcast episode be.

Within an episode, a jump forward in time needs its own audible signal, because the listener has no page break to see. See how to signal a time skip in an audio drama.

Why AI voices make length cheap to experiment with

With a traditional cast, length is expensive to change. A tighter cut means booking studio time and re-recording lines, so most producers lock a runtime early and defend it. That pressure is why so many first episodes are either bloated or clipped: nobody could afford to test the alternative.

When every character is an AI voice, that constraint largely disappears. Trimming a scene, splitting a long episode into two, or stitching two short ones together is a matter of editing the script and re-rendering, not re-hiring anyone. You can produce a 12-minute cut and a 20-minute cut of the same episode and actually listen to both before you decide. The only real cost is your monthly word allowance, so you can iterate on pacing the way a writer iterates on a draft, rather than committing to a runtime on paper and hoping it plays.

That changes the workflow. Instead of guessing the right length, you write to a target, render it, listen on the device your audience uses, and cut or expand from there. Length becomes something you tune by ear, which is exactly how it should be decided.

Turning a length target into a finished episode

AudioProducer.ai is built to make that loop fast. You paste your script or import an EPUB, and Auto-Assign tags every line by speaker so each character gets a distinct voice. Auto-Assign Sounds places music beds, ambient soundscapes, and one-shot effects to match the scene, and you review and adjust all of it in the editor. One click renders the whole thing into a finished audio file you download; there is no separate editing software to buy.

Because episodes map to chapters in a project, changing an episode's length is just changing how you break the script: merge two chapters for a longer feature, or split one for a tighter serialized cut, then re-render. You can proof-listen a cut, trim the scenes that drag, and render again until the pacing lands, all within your plan's monthly words. The output is a standard audio file you own and can upload wherever you publish; AudioProducer produces the episode, it does not distribute it for you.

Try it free and render a first episode to hear where your length actually wants to land. Browse more guides on the AudioProducer blog.