What If Authors Owned Their Readers Again?

This is not a description of a finished product. This is a description of the platform we want to build, and the principles we want to build it under. Some of it exists today. Most of it does not. We are writing it down now so the work has a target to walk toward — and so anyone considering us can hold us to it.

The problem we keep walking into

It is 2026. You wrote a book. You want to publish it as an audiobook, because audio is now half the book market and growing.

Your options:

You can go to ACX (Audible/Amazon). They keep 60% of every sale. They lock you into seven years of exclusivity. Their policy on AI narration shifted three times in the last eighteen months and may shift again. To be accepted as a narrator, your audio needs to clear their QC. If you do not pass, you hire someone. A professional narrator costs £200–400 per finished hour. A six-hour book is £1,200–2,400 before you sell a single copy.

You can publish text-only on Kindle Direct. The built-in TTS is robotic and most readers turn it off. No serious reader experiences your book as audio.

You can pay ElevenLabs or Speechki to generate AI narration. Beautiful audio. Then what? They generate the file. They do not sell it for you. They do not host it. They do not give your readers a way to listen with sync to the text. They do not let your readers chat with your characters. You walk away with an .mp3 and you are back to "where do I sell this."

You can DIY: ElevenLabs for voice, Calibre for conversion, Gumroad for sales, your own website for a player, some other tool for sync, a fourth tool for community. Five subscriptions. Five logins. Five places where something can break.

And the reader? The person who actually wants to read and listen to your book?

They can buy your text on Kindle, your audio on Audible, and use Amazon's WhisperSync — but only if both versions came through Amazon's pipeline, which most indie books do not. They can use Storytel for both modes but Storytel does not sync them. They can listen on Spotify Audiobooks, which has even worse terms than Audible.

Both sides lose. The author gives up 60% of every sale and seven years of rights to reach an audience. The reader pays twice or accepts a fractured experience.

This is the situation we are building MimicReader to replace.

What we are building

A single platform where an author publishes once, and the reader gets the whole book — text, audio, character chat, voice cloning, sync between modes — for one price.

For the author

You upload your manuscript. EPUB, TXT, PDF, MOBI, FB2 — whatever you have. We parse chapters, detect dialogues, normalize the text.

You record five seconds of your own voice on your phone. We clone it. Now your entire book can be narrated in your voice — or you choose from a library of voices in 23 languages. You pick the quality tier that fits your book. You hit generate. The book renders in a few hours on our GPU.

You upload a cover. You write a description. You set your price.

Your book goes live on a public page at mimicreader.ai/author/your-name/your-book-title. It includes a player, a sample chapter, a sync between text and audio, and a chat box where readers can talk to characters from your book in your book's voice.

You can share this page anywhere: Substack, Twitter, your own website, your mailing list, Reddit. There is no walled garden. There is no exclusivity. You can simultaneously sell the same book on Gumroad, Payhip, ACX, anywhere. We do not lock your manuscript, your audio, or your audience. Pricing details will be published openly when the marketplace opens; the principle is simple — the author keeps the lion’s share, and we never grow our cut as the business grows.

For the reader

You browse books. Some are public domain classics, free. Some are by independent authors, priced by the author. You buy one.

The book opens in our reader. You can:

• Read it as text, paginated like a real book
• Listen to it as audio
• Read along with the audio, paragraphs highlighting as they are spoken
• Chat with characters — ask Aslan a question, argue with Mr. Darcy, ask the narrator why a chapter ended that way
• Switch devices and your progress follows you
• Adjust speed, switch voice, change language if the author offered translations

You bought the book once. It is yours forever. If the author leaves the platform, your book stays in your library. We do not yank books out of buyers' libraries — that breaks the basic promise that you can buy something and own it.

How the economics will work

Most platforms hide their economics. Ours will be on a public page that any author can read before signing up, in plain English, no asterisks.

The principles we will not move on, no matter how big the business gets:

• The author keeps the lion’s share of every sale. Our cut covers payment processing and the platform itself, nothing more. We never raise it.
• A modest, transparent platform subscription covers the cost of keeping a book online, the AI Writing Studio, voice cloning, and the analytics dashboard. One number, on the same page as everything else.
• No exclusivity. No price markup. We sell at the price the author sets. We do not bundle their work into our own subscriptions without consent.
• The marketplace is a service to authors, not a revenue stream for us. Our income should grow because more authors choose us — not because we extract more from each one.

For context, here is what the market does today:

As far as we know, no one else is trying to put all of this in one place — for one transparent price the author actually understands.

What this means for writers

You stop choosing between "give Amazon seven years" and "do everything yourself with five tools."

You publish once. You distribute everywhere — including back on your own site, your Substack, Gumroad, whatever — because we do not ask for exclusivity.

You earn the lion’s share of every direct sale through MimicReader, on terms we publish openly and do not move.

You give your readers something Audible cannot: a unified text-and-audio experience with characters they can interact with.

You can pivot, leave, take your manuscript elsewhere any day. Cancel your platform subscription and your book remains in the libraries of everyone who already bought it, indefinitely. Your account goes dormant. New buyers see "this author is inactive" and any royalty queues up for if you ever come back.

You are not trapped. That is the point.

What this means for readers

You buy a book. You get both modes. Text and audio synced. Characters you can talk to. Your highlights, bookmarks, and progress sync across your devices.

You pay the author directly, not through three middlemen.

The books you bought stay yours. We do not have a switch that removes them from your library because someone above us decided to delist a title.

You can read on your phone, listen on a walk, talk to a character on the bus. Same book.

You pay what the author asks. We add nothing on top. We do not markup books, we do not tier-gate features, we do not show you ads.

Why we are building it this way

Three observations drive everything.

1. TTS is becoming a commodity. Web Speech API in your browser is free. Google's Gemini Flash TTS is free. On-device TTS via Whisper variants is shipping on phones now. In two years, "generate audio from text" will be a free button on your operating system. Anyone whose business depends on charging for "AI generated audio" is selling a feature that will be free.

The moat is not the audio engine. The moat is the platform — the marketplace, the reader, the rights model, the trust.

2. Audible's dominance depends on exclusivity, not quality. The reason an indie author tolerates 60% taken is because Audible has the readers and the recommendation algorithm. Take that away — let authors keep their audiences direct via their existing channels — and the 60% becomes indefensible. Substack proved this works for text. We think it works for audio too.

3. Apple and Audible accepted AI narration in 2025. That was the final wall. AI-narrated audiobooks are now mainstream. The cultural objection — "but a human should read it" — collapsed when the biggest distributors blessed it. The technical objection — "AI sounds robotic" — collapsed when models like ours and ElevenLabs surpassed mid-tier human narrators.

These three together create an opening: the platform that helps indie authors publish audio cheap, keep their audience, and give readers a better experience than Audible can.

That platform is what we are building.

Where we are today (honestly)

We are not there yet.

As of May 2026 — about five weeks after our official launch on UNESCO World Book Day (April 23) — we have:

• 47 registered users
• 1 user has uploaded their own book
• 4 users have actively listened to generated audio
• 0 authors have published a book through our marketplace
• 0 marketplace transactions have happened, because the marketplace does not exist yet

What does exist and work today

• Audio generation in 23 languages (OmniVoice and Chatterbox engines on our own GPU + RunPod overflow)
• Voice cloning from a 5-second sample
• Read-along sync between text and audio, with paragraph highlighting and karaoke-style word timing
• Chat with characters of any uploaded book
• The reader app with paginated reading, bookmarks, progress sync
• A free Live Reader tier using Gemini for users who want to try without paying
• Multi-language translation (23 languages) using a stack of free + paid engines
• A children's learning module with phonics curriculum and AI tutor
• A working payment integration for users who want to pre-load credits
• 11 languages of UI translation, auto-detect from browser
• The whole thing live at mimicreader.ai

What does not yet exist

• A marketplace with public author pages and book pages
• The author-side platform subscription (the foundation of how we will sustain ourselves without taxing every sale)
• Direct payouts to authors via Stripe Connect
• Author verification, fraud protection, DMCA flow
• Sales analytics, royalty dashboards, payout management
• Reader subscriptions for the whole library

Our next 6–12 weeks are about closing the gap on the toolkit (the .m4b master, metadata, cover, public preview page, download bundle) — what we call internally Phase A. After that, Phase B is the actual marketplace: Stripe Connect, public author pages, payouts. That is likely a 3–6 month project.

We are writing this vision down now because we believe in it, not because we have shipped it. The version of MimicReader described above is the one we are committing to build.

What you can do today

If you are an author considering whether MimicReader is worth your attention: today you can sign up free, upload your manuscript, generate audio with our free starting credit, and clone your own voice in 5 seconds. There is no marketplace yet — you cannot sell through us yet — but you can see whether the tooling holds up. We would love to hear what is missing.

If you are a reader: today you have access to 100,000+ public domain books with audio, a free Live Reader, and the ability to upload your own files and have them read. No commitment.

If you are a developer or fellow founder: we are solo and we are transparent. The fact that this essay exists on our public blog — before we have a marketplace, before we have proof — is part of that commitment. We would rather be held to a written vision than quietly drift away from it.

In the next year, our success will be measured by one thing: how many indie authors chose to publish through us, kept the lion’s share of their revenue, and felt better off doing so than going through Audible.

If that number is meaningful, we built something useful. If it is zero, this essay was just words and we learned something.

We are writing the future as we are building it. Both are in motion.

Our commitments — the only fixed points

These will not change as we build, ship, pivot, or grow:

If you ever see us break one of these six, call us on it.

— Sebastian
Founder of MimicReader. Solo. Working on this from Scotland. Reachable at [email protected].

This document is a draft of what we are building. The next revision will follow our first real conversation with an author who tried to publish through us.