Questions, answered.

A gaming PC or gaming laptop with an 8 GB NVIDIA GPU is enough — that's the benchmarked path, generating at roughly real-time. Apple-silicon Macs are supported too: the engines run on the Mac's GPU (or CPU), just slower. No GPU at all works, but it's slow enough that we'd call it a fallback, not a plan. AMD GPUs are an experimental preview: Qwen and Coqui run on ROCm and Kokoro on the CPU, with an automatic CPU fallback if the ROCm install doesn't take — a ROCm-supported card and a recent driver are needed. Analysis — the character-detection step — runs on a local CPU model or the free Gemini tier either way, so the GPU stays free for speech.

Two ways, both free. One click: if you use Pinokio, paste the Castwright repo URL, click Install, then Start — Pinokio brings its own Python and tools, so there's no terminal and nothing to set up. Or direct: download the zip from the download page, unzip it, and run the launcher; one build runs on Windows, macOS, and Linux. Either way the first launch sets up your voices. See /docs/install for the full walkthrough.

Yes. Your books, voices, and library stay on your machine, and synthesis runs entirely locally once models are installed — the generated audio never leaves the house. One scoped exception, stated plainly: the optional cloud analyzer (Gemini free tier) sends chapter text for character detection. Prefer zero exceptions? Point Castwright at a local Ollama model and the whole pipeline runs offline.

For personal use of books you already own, format-shifting for accessibility is generally accepted in most jurisdictions. Castwright is a personal-use tool — it is not designed or intended for redistribution of audiobook copies. Respect copyright and the terms of your ebook purchase; we say so plainly rather than burying it.

Import: EPUB, PDF, or paste plain text directly. Export: M4B (chapter-marked audiobook, works in every major player), MP3 (flat audio, maximum compatibility). Your files land on your drive — no upload required.

Yes, two ways. Today, free: export M4B or MP3 and drop it into any player you like. At launch, the companion app (iOS and Android) joins the Cast Pass — pair with a QR code, sync your library over your own Wi-Fi, download books, and listen offline with chapters, speed control, and your place kept in sync.

Source-available, not OSI open source, and we say so plainly. The code is public on GitHub, under the Functional Source License (FSL-1.1-Apache-2.0): fully readable, forkable for personal use, and it converts to Apache-2.0 two years after each release. Competing redistribution is barred during that window. See /legal/licensing for the full picture.

Yes, once models are installed. Generation and playback are fully offline — nothing phones home during synthesis. The optional cloud analyzer needs a connection; if you configure Ollama as the analyzer instead, the whole pipeline runs without internet access.

Nothing, during the public beta: it's free for every book — the full cast, series memory included, yours to keep. At launch, the Cast Pass — seven dollars, once — unlocks the cross-book series memory and the companion app. No subscription, no metering. Free buys you a perfect book; the Cast Pass gives your library a memory. See /pricing for the breakdown.

Yes — voice cloning is coming in the next release. You record a short consent clip, it stays on your machine, and Castwright can narrate in your voice. The cloned voice is never uploaded or shared. Your consent recording is yours to keep or delete.