Connect Kavita, Komga, Calibre-Web or Audiobookshelf — or any OPDS server. Browse your books on your phone. Generate AI audiobooks in 23 languages. Listen offline.
Add your serverKavita, Komga, Calibre-Web, Audiobookshelf — great servers. Their mobile story is the gap.
Kavita, Komga and Calibre-Web are web-only. Their interface is clunky on phones. Tachiyomi/Mihon helps for manga, but novels need a real reader.
None of them can turn your books into audiobooks. System TTS sounds robotic. Audiobookshelf only plays existing audio — doesn't generate.
Most need a constant connection back to your server. On the underground, on a flight, in the countryside — your library disappears.
One mobile companion for every OPDS-compatible ebook server you self-host.
Beautiful mobile reader as a PWA. Install on your home screen. Paginated reading, adjustable fonts, dark mode, chapter navigation. "Save offline" works on flights, in the underground, or anywhere your server isn't reachable.
Pick a book from your Kavita/Komga/CW library → tap Generate audiobook. AI narration with emotion detection in 23 languages. Download as M4A — yours forever, no DRM.
Standards-based OPDS protocol. Same setup flow whether you run Kavita, Komga, Calibre-Web, Audiobookshelf, or any other server that exposes an OPDS feed.
Your library stays on your server. MimicReader fetches a book only when you actively import it. No silent background uploads, no double-storage. Catalog browse is instant.
Credentials encrypted at rest (Fernet/AES-128). HTTPS-only. SSRF protection blocks private and metadata IP ranges. We never log your password or content. You can delete the connection anytime.
Read your library in any of 23 languages. Free Helsinki MT + open-source Ollama, or premium Claude Sonnet for publisher-grade quality. Original stays untouched.
Same three-step setup for every supported server.
API key from Kavita / Audiobookshelf, or username + password for Komga / Calibre-Web. Takes 30 seconds — instructions per platform below.
Open Dashboard → Connections. Pick server type, paste URL + credentials. We test the connection and store the credentials encrypted.
Your library appears in MimicReader. Drill into categories, see covers, click any book to import. Read in our reader, generate an audiobook, or download offline.
Don't have a server yet? Try with the public Komga demo — one click adds it for testing.
Click the platform you self-host. Universal flow, platform-specific credential location.
Web-based ebook/manga server. Excellent for personal libraries but no native mobile app — "mobile app" is the most-requested feature on their GitHub.
Manga / comic / ebook server. Supports OPDS v1.2. Pairs nicely with Mihon for manga, but novels and AI audio need MimicReader.
https://komga.example.com) + Komga username + password./opds/v1.2/catalog.Web wrapper around your Calibre library. Built-in OPDS feed at /opds. Great for sharing books across a household.
/opds.Self-hosted audiobook + ebook server. Has its own mobile apps for playback, but doesn't generate audio — that's where MimicReader fills the gap.
Run a different ebook server? If it exposes any OPDS catalog (Atom feed at /opds or similar) MimicReader can browse it. Pick Generic OPDS, paste the catalog URL, choose Basic / API key auth.
Running Calibre desktop instead? It needs a different setup — we have a free plugin: MimicReader for Calibre →
No other tool combines mobile browsing of your self-hosted library with AI audiobook generation.
| Feature | MimicReader | Kavita web | Komga web | Calibre-Web | Audiobookshelf |
|---|---|---|---|---|---|
| Native mobile reader (PWA) | Full PWA, install prompt | Web-only, basic PWA | Web-only | Responsive web | Native app + PWA |
| Generate AI audiobooks from any book | 23 languages, neural TTS | No | No | No | Plays only, doesn't generate |
| Voice cloning | From 5s sample | No | No | No | No |
| Offline reading + audio | Books + audiobooks cached | Needs server | Needs server | Needs server | Audio only |
| Translation (any ↔ any) | 23 languages | No | No | No | No |
| Chat with book characters (AI) | Built-in | No | No | No | No |
| Karaoke read-along (synced word highlight) | Built-in | No | No | No | Sync with existing audio only |
| OPDS connector (browse remote) | Universal: Kavita / Komga / CW / ABS | Self only | Self only | Self only | Self only |
| Self-hostable | Cloud (your account) | Yes | Yes | Yes | Yes |
| Price | Free reader + PAYG audio | Free, self-host | Free, self-host | Free, self-host | Free, self-host |
We're not a replacement for Kavita/Komga/CW/ABS — we're the mobile companion + AI audio layer they don't have.
No installation. No port forwarding. Just paste your server URL.
Credentials encrypted at rest. HTTPS-only. SSRF protection built-in. We never see your books unless you import them.
Kavita, Komga, Calibre-Web, Audiobookshelf — tested. Plus any other server that exposes an OPDS feed (Atom XML catalog standardised in 2010). If your server has /opds, MimicReader can probably read it.
No. Kavita, Komga, Calibre-Web and Audiobookshelf already expose an OPDS endpoint out of the box. You just paste the URL + credentials in MimicReader's dashboard. Nothing is installed, nothing is modified.
Yes. We encrypt them at rest using Fernet (AES-128 + HMAC-SHA256), with a key derived from our server-side secret. Credentials are never logged. The connection is HTTPS-only with strict TLS validation, and SSRF protection blocks attempts to reach private/internal IPs (RFC1918, loopback, AWS metadata) on every redirect. Body size is hard-capped (10 MB feed, 100 MB book) so a malicious server can't DoS our backend.
It needs to be reachable from MimicReader's backend — which means reachable from the public internet. If you currently keep it on LAN only, the easiest options are: Cloudflare Tunnel (free, no port forwarding), Tailscale Funnel, or a reverse proxy with HTTPS. We recommend a real certificate (Let's Encrypt) over self-signed.
Kavita and Komga have web UIs, but they aren't built mobile-first — they're built around a desktop reading flow. MimicReader is a true PWA: install icon, paginated touch reading, dark-mode optimised, offline cache, and most importantly — you can generate an audiobook from any book in your library. None of those servers have AI TTS.
The PWA caches the books you've imported and the audiobooks you've generated. Once cached, you read & listen without a connection. Browsing the remote OPDS catalog requires the connection because it's live data from your server.
Reading is unlimited and free. The OPDS connector is free. Every account gets 1 free hour of AI audiobook generation per month. Beyond that it's pay-as-you-go from £1 per hour of finished audio. Credits never expire.
Yes — that's the headline feature. After importing a book through the OPDS connector, tap Generate audiobook. AI narration with emotion detection in 23 languages. The finished M4A is yours forever — download it, no DRM, no lock-in.
We use Chatterbox + OmniVoice neural TTS engines. They detect dialogue, emotion and pacing — no flat robotic voice. In blind tests our voices are preferred over ElevenLabs 63% of the time. Voice cloning from a 5-second audio sample is also available on paid tiers.
Yes. Dashboard → Connections → Delete. Books you've already imported stay in your MimicReader library. The connection (URL + encrypted credentials) is removed from our database immediately.