Skip to main content
November Sale: 30% OFF generation »

Audible Now Pitches Audiobooks Out Loud. The Sample Is the Part You Made

Audible's new audio previews pair an AI-voiced overview with the book's own narrated sample. The overview is the store's. The opening minutes are yours.

Midsummerr||6 min read
Watercolor icon of an open book with a pair of headphones resting on its pages
Hear a full production first

Audio Sample

Hear a full cast before you read on

Scenes from published Midsummerr productions: every character voiced, scored and sound-designed. Judge the format for yourself.

On November 16, Audible started pitching audiobooks out loud. Its new feature, audio discovery, plays a short spoken preview before a listener commits to a title: a brief overview of the book's themes and what listeners like, commentary from the author or an Audible editor, and the audiobook's own narrated sample. The overview is read by Audible's virtual voice, which the company describes as an AI-generated speech technology. It starts on select English-language titles in the US.

Our reading: the shop window for audiobooks has moved from the page to the ear, and inside that window the store now describes the book in its own voice. The one part of the preview that comes from the book's own production is the sample. ACX asks for that sample to be the opening five minutes, and those minutes now carry the pitch.

What Audible launched

Audible's announcement lists where each part of the preview comes from. Laid side by side, it shows who made each part.

Part of the previewWhere it comes fromWho made it
Overview of themes and what listeners likePublicly available synopses, metadata and reviews, read by Audible's virtual voiceThe store, from your material and your readers' reviews
CommentaryRecordings of past Audible interviews with the author, or an Audible editorAudible's archive
Narrated sampleThe audiobook itselfThe book's production

Audible says the previews can be heard on the go and right after a listener finishes a title, the moment they are choosing what comes next. "Picking your next listen should feel like getting a tip from a friend," said Andy Tsao, Audible's chief product and analytics officer.

The same release adds collections personalized by listening taste, which Audible says are meant to help listeners look beyond the bestseller lists. Two tools from earlier this year are expanding too. Recaps gives a spoiler-free audio summary up to the point where the listener paused, and Ask a Question answers questions about a character, a plot point or a historical reference. Both launched in beta on select Audible Originals and public domain titles, are now reaching more US titles, and begin rolling out broadly in December, for listeners who own the title or have it through their membership.

Ready to try it on your own book?

Start with 5,000words free →

The description is now the store's

For years a listener choosing an audiobook read: a cover, a blurb, a star rating, a few reviews. Audio discovery turns that into something heard, and it hands the telling to the store. The overview is built from public material and read by the same speech technology for every title in the program. TechCrunch noted one risk worth watching: if the summary draws mostly on positive remarks, negative reviews could be underrepresented. Audible has not said how the overview weighs them.

Two consequences follow for an author.

Your synopsis is now a script. The overview draws on your synopsis and metadata. A blurb written for the eye, built on a teaser question and a list of comparable titles, gives a spoken summary less to work with than one that names the premise, the central relationship and the themes in plain sentences. We made a similar argument about Spotify's conversational search, where the description became the thing listeners query. On Audible, it also becomes the thing they hear.

What listeners like is now read aloud. The overview includes a summary of what listeners like, drawn from public reviews. Reviews that mention the performance, a voice that fit a character or a scene that landed, are material that summary can draw on. Audible has not published how it chooses which reactions to include.

The opening five minutes are the proof

Everything else in the preview is about the book. The sample is the book. ACX's audio submission requirements cap the retail sample at five minutes and ask submitters to "choose the first five minutes" of the audiobook, so the sample a listener hears in audio discovery is, as a rule, the opening.

Five minutes is enough for a listener to decide three things: whether they like the voice telling the story, whether the characters sound like people, and whether the book sounds like a world or a reading. Here is how to make those minutes count.

  • Hear the opening as a stranger would. Play the first five minutes without the book in front of you. If a listener could not tell who is speaking, or nobody speaks at all, that is the sample a stranger gets.
  • Get the first scene's voices right before anything else. The characters who speak in the opening are the ones a listener judges the whole cast by. Audition them against that scene first.
  • Check the opening for explicit material. ACX does not allow it in a sample; if your first five minutes contain it, you will need a different excerpt, and it should be one you chose on purpose.

The audiobook file requirements guide covers the rest of a retail delivery, from credits files to loudness.

Where this applies if you produce with AI

Audio discovery runs on Audible, and Audible's narration rules have not changed with it. ACX's requirements say a submitted audiobook "must be narrated by a human unless otherwise authorized", and Audible says it is working to accept third-party text-to-speech content from publishers and creators, with timing to come. For an author producing with AI today, this preview is a view of where audiobook stores are heading rather than a shelf to sit on. The lesson travels: wherever a listener tries before they commit, the opening minutes are what they try.

In a Full Production on Midsummerr, the opening is produced like the rest of the book: a narrator and a voice for each speaking part, cast from how the book describes them, with an original score and sound design. Martin, the AI audio director in the studio, fine-tunes the opening scene by note. Tell it "make her first line colder" and it records a new take of that line without regenerating the chapter, and keeps the earlier take in case you want it back. Midsummerr is the only audiobook platform with an AI audio director, and how to change one line of your audiobook shows the whole flow. To hear what a produced opening sounds like, The Strange Case of Dr. Jekyll and Mr. Hyde streams free: Hyde's theme enters in the first chapter, on the first line that speaks his name. To try it on your own book, see what a production includes; the first 5,000 words are free.

The honest version

This is a launch announcement, not a result. Audible has not said how many titles have audio discovery, which ones, or whether authors and publishers can choose to be in or out of it; the release says the features were developed with input from creators and that Audible welcomes creator feedback. It does not give the order in which the three parts play. And nobody outside Audible has measured yet whether a spoken preview changes what listeners pick.

What is settled is the direction. Amazon's Audible now introduces books with its own AI-generated voice, describes them from material the author wrote and readers posted, and lets the book's own audio speak for itself in the sample. The description is shared. The opening is yours.

Further reading

Key takeaways

  • Audible launched audio discovery on November 16, 2026: a spoken preview that combines the narrated sample with an overview of the book's themes and what listeners like, plus author or editor commentary. Select English-language US titles first, with plans to expand.
  • The overview is read by Audible's virtual voice, an AI-generated speech technology, and draws on publicly available synopses, metadata and reviews. Author commentary comes from recordings of past Audible interviews.
  • Recaps and Ask a Question, launched in beta earlier this year, are expanding to more US titles and begin rolling out broadly in December, for listeners who own the title or have it through their membership.
  • The store now describes each book in its own voice from the same kinds of public material. The narrated sample is the only part of the preview made by the book's own production, and ACX asks for the first five minutes.
  • Treat the opening five minutes as the pitch: a speaking character early, voices that match the page, and a synopsis written to be read aloud, because it is now source material for a spoken summary.

See where your book fits

Dramatized audio is the format pulling listeners. Put a chapter through and judge the result yourself.

Keep reading