These run locally on your server. No API keys needed, no data leaves your machine.
Text-to-Speech (Kokoro): checking...
Speech-to-Text (Whisper): checking...
Text-to-Speech Voice
Choose the voice used when generating audiobooks from ebook text. All voices are powered by Kokoro and run locally.
Speech-to-Text Model
Choose the Whisper model used when transcribing audiobooks to text. Larger models are more accurate but slower. Changes apply to new transcription jobs.
The Whisper model uses ~3 GB of RAM (or VRAM on GPU). Unloading frees memory when not actively transcribing.
Book Q&A
Add your own AI API key to enable intelligent question and answer about your books. Your key is stored locally on this server and only sent to the provider you select.
Voice Chat
Talk to your books using real-time voice conversation. Requires a speech-to-speech API key. This feature sends audio to an external service.
Voice chat is a premium feature that uses real-time speech-to-speech AI. Your audio is sent to the selected provider for processing.
Mobile App Pairing
Open the abookify app on your phone and scan this QR code,
or enter the server URL manually.