MimikaStudio

Why we included this project

Voice cloning and TTS tools often force you to choose between a desktop app and a server you can script. MimikaStudio tries to be both, and it runs entirely on-device on Apple Silicon. You can clone a voice from a few seconds of reference audio with Qwen3-TTS or Chatterbox, or generate speech with the faster Kokoro and Supertonic models, all behind one interface. It also works as a document reader: it reads PDF, DOCX, EPUB, Markdown, and TXT files aloud with sentence-level highlighting, and can queue up a whole book as an audiobook using voice presets you can reuse. If you'd rather drive TTS from your own scripts, the same engine exposes an agentic server with a job queue and an API. One caveat: binaries are macOS-only for now, so Windows users will have to wait for the planned release.