The problem.
There’s barely any audiobook selection in european portuguese, my native language. i do have a solid collection of ebooks sitting on my disk though, and i’d love to listen to them the same way i listen to podcasts — except there isn’t really a decent audiobook app on desktop either.
my solution.
Capybara is a single desktop app that does two things: it generates an audiobook from any EPUB, PDF, or plain text file using neural TTS voices, and it plays the result back with proper chapter support, persistent progress, sleep timers, and adjustable speed.
It’s built with Tauri, Svelte 5, and Rust. The synthesis pipeline runs concurrently with dynamic rate-limit throttling (AIMD, like TCP congestion control) so long books don’t take forever and don’t get the connection killed.
Still a work in progress. Proper release post will follow once it’s ready.