EbookAloud, a new tool that launched on Hacker News this week, turns ebook files into finished audiobooks using AI voices for between $0.99 and $5.99 per book. There is no subscription and no account required: you upload a file, pick a narrator, preview the chapters, and pay only for the conversion you keep.
What Happened
Built by Trilunar Labs, EbookAloud accepts EPUB, DOCX, HTML, Markdown, and TXT files and outputs an M4B audiobook, the chaptered format that Apple Books and other players expect. The service runs on Kokoro-82M, a small open-weight text-to-speech model released under an Apache license that has become a favorite for long-form narration. EbookAloud offers seven preset voices, including Heart, Bella, Nicole, Adam, Michael, Emma, and George, and lets you choose whether to include front and back matter. A free sample conversion of Alice's Adventures in Wonderland lets you hear the output quality before paying.
Why It Matters
Self-published authors have long faced a wall when it comes to audiobooks: human narration can cost thousands of dollars per title. AI narration collapses that to a few dollars, which is why platforms keep racing into the space. Spotify recently rolled out an ElevenLabs-powered audiobook tool, covered in our piece on Spotify's audiobook creation tool. EbookAloud takes the opposite tack from those platform plays: it is unbundled, pay-per-book, and exports a standard file you can load into Apple Books or any player you already use, rather than locking the result inside one storefront.
Key Details
Pricing scales with length: up to 20,000 words is $0.99, up to 50,000 is $2.49, up to 100,000 is $3.99, and up to 300,000 words is $5.99. Finished files stay available for 48 hours after processing, so you download and keep your own copy. EbookAloud is essentially a hosted, no-setup version of the do-it-yourself route: the open-source audiblez command-line tool runs the same Kokoro pipeline locally and for free, but requires Python, ffmpeg, and the patience to configure them. The output of either approach drops cleanly into self-hosted libraries like AudioBookshelf. EPUB produces the best results because chapter structure is detected automatically; PDF support is listed as coming soon.
What to Do Next
Run the free Alice in Wonderland sample first to judge whether the Kokoro voices fit your material, since AI narration still struggles with heavy dialogue and unusual names. If you write nonfiction or clean prose, the result is usually strong enough to publish. Export your manuscript to EPUB rather than DOCX or PDF for the cleanest chapter splits, and budget a listen-through pass to catch any mispronounced proper nouns before you distribute. For authors weighing a hosted converter against running the pipeline yourself, the choice comes down to whether you would rather pay a few dollars to skip setup or invest an afternoon configuring a local toolchain you can reuse for free.
