Two of the five big audiobook storefronts will take a finished file you narrated yourself with a synthetic voice. The other three will give you AI narration, but only their own, generated from your ebook inside their pipeline, with their voices.
That single distinction is the whole planning problem. Most guides answer "does platform X allow AI narration" with a yes or a no, which is useless once you are holding 90 minutes of audio you already mastered. The question that matters is whether the store accepts a file or only a nomination.
The platform matrix at a glance
| Platform | Accepts AI narration | Disclosure mechanism | Notes |
|---|---|---|---|
| Spotify for Authors | Yes, including audio you produced yourself | Upload option "This audiobook uses digital voice narration"; Spotify adds a sentence to your description | Not currently shared to referral partners |
| Kobo Writing Life | Yes, including audio you produced yourself | Narrator contributor listed as "Synthesised voice - male / female / unspecified" | At least two contributors required, one author and one narrator |
| Google Play Books | Yes, generated by Play Books from your EPUB | Built into the auto-narrated program | Beta, no program fee; price parity rules apply if you sell the downloaded file elsewhere |
| Apple Books | Yes, generated by Apple from your ebook | Store credit reads "Narrated by Apple Books" | Delivered via preferred partners only; English; one to two months to process |
| Audible via KDP | Yes, invite-only beta using Amazon's virtual voices | Handled inside the program | US marketplace only at present |
| ACX standard upload | Check the current requirements directly | Not applicable | See the ACX help centre for current submission rules |
Everything below comes from those policy pages, read the same week this went up. Policies here move, so treat this as a map of the doors rather than a permanent floor plan.
Can you publish an AI narrated audiobook on Spotify?
Yes, and Spotify is the least ambiguous of the group. The Spotify for Authors help page on digital voice narration says "We accept audiobooks with digital voice narration on Spotify." At upload you select the option This audiobook uses digital voice narration, and then, in Spotify's words, "We'll add a short sentence to your book's description to let listeners know it uses digital narration."
Two details are worth internalising. The disclosure is reversible: replace the synthetic narration with a human recording later and you unselect the option, and Spotify removes the sentence. The cost of the toggle is reach, because Spotify notes: "We don't currently share audiobooks with digital voice narration to referral partners." Spotify names Google Play Books and ElevenLabs among accepted providers, "and more", which is why a self-produced file passes rather than only vendor output.
Does Google Play Books accept auto narration?
Google accepts AI narration by making it for you. The auto-narrated audiobooks program policies describe a flow that starts with an .epub in Partner Center, moves through voice choice and narration edits, then publishes to the Play Store. You must own the audiobook rights to any title you auto-narrate. During the beta there is no program fee.
The clause creators skip is the pricing one. You can download the finished audiobook and sell it on other platforms, but it must also be for sale on Google Play in every country where it is available elsewhere, and the Play list price must not be higher than the price on those other retailers. Google sets previews on a scale too: one minute for runtimes of ten minutes or less, thirty minutes for runtimes of 300 minutes or more, and up to ten percent in between.
Does Apple Books accept AI narration?
Apple accepts it, produces it, and brands it. The digital narration pages describe speech synthesis worked over by linguists, quality control specialists and audio engineers, and the getting started page states that digitally narrated titles appear in the store as "Narrated by Apple Books". Production is free.
The constraints are real, though. You cannot deliver directly: nominations go through preferred partners, currently Draft2Digital, Ingram CoreSource and PublishDrive. The ebook must already be on Apple Books, you must own the audio rights, and the book must be in English. Nominations get judged on file quality, content compatibility and editorial review, with complex formatting, heavy foreign-language vocabulary, poetry, cookbooks and graphic novels called out as poor candidates. Apple is direct about the rest: even a title that meets every requirement may not pass synthesis. Processing runs one to two months, and pre-orders are not supported. You keep the audiobook rights and can still produce your own edition elsewhere.
Does Kobo accept AI narration?
Kobo is the second door that takes your own file, and it is refreshingly plain about it. The Kobo Writing Life help article on AI narration and audiobooks says Rakuten Kobo gladly accepts the use of AI narration, and asks for at least two contributors, one author and one narrator. When the narration is synthetic, the narrator contributor goes in as "Synthesised voice - male", "Synthesised voice - female" or "Synthesised voice - unspecified".
That is the entire disclosure mechanism: a metadata field. Which also means it is the easiest one to fumble, because nothing stops you typing a fictional narrator name into it.
Does Audible accept AI narration in 2026?
Partly, and through a narrow door. Amazon's KDP help centre describes Audiobooks with virtual voice as an invite-only beta for eligible KDP ebooks. If your ebook qualifies, an "Add Audible audiobook with virtual voice" option appears on your Bookshelf. There are more than 50 virtual voices, you can set a different voice per chapter, you can edit the audio in the Virtual Voice Studio, and it is free at this time. Distribution is limited to the US marketplace.
For narration you produced yourself, the ACX submission path is the one to verify by hand rather than trust a blog post about. The ACX help centre did not render when we tried to read it for this piece, so check the current requirements directly at help.acx.com before you commit to a production plan built around it. Our ACX compliance guide goes deeper on the technical specification and the eligibility question; this post is the multi-platform decision that sits upstream of it.
Why the biggest channel still shapes your plan
Add it up and a pattern falls out. The stores where you control the voice, the performance and the master are Spotify and Kobo. The stores with the largest audiobook audiences hand you a catalogue of their voices instead. If your reason for going synthetic is craft, a specific voice and pace that stays consistent across eight books, the platforms that reward that choice are not the ones with the biggest storefront.
So plan distribution before production, not after. Pick the two or three destinations you actually want, read their current policy pages yourself, then produce once to the strictest technical target among them. Producing first and shopping afterwards is how creators end up with a mastered file that only one retailer will take.
A production order that respects the policies
Start with the destination list. Then confirm the disclosure field on each one, because the wording is fixed and you will need it in your metadata sheet anyway. Then narrate, master, and export to the tightest specification in your set.
That order is what the audiobook production kit is built around: chapter structure, a pronunciation pass, mastering targets, and a metadata checklist you fill in before you record rather than after. If the decision is still cost against a human narrator, the AI voice cost calculator will do that arithmetic with your own chapter count and runtime, and the pricing page has the current numbers for producing unlimited audio at one flat monthly price.
Two more posts worth reading alongside this one: the manuscript to audiobook workflow for the actual production sequence, and audiobook narration tips for the direction choices that make a two-hour listen bearable.
What to write in your disclosure
Keep it boring and accurate. On Spotify you do not write anything; you tick the box and Spotify writes the sentence. On Kobo you write the exact contributor label. Where a store gives you free text, one line does the job: "This audiobook was narrated with a synthetic voice."
Nobody has ever refunded an audiobook because the description was honest. They refund it because the narration surprised them.
The Vois Team