PowerBeef/Vocello

Vocello: a local, private voice studio for Apple Silicon. Write a script, pick or describe a voice, and generate speech on-device, faster than realtime on an 8 GB M2 Mac mini. Native Swift + MLX, no Python. Mac app out now, iPhone beta on TestFlight. (Formerly QwenVoice.)

What it solves

Vocello is a local-first, private voice studio that allows users to generate high-quality speech from text on Apple devices. It eliminates the need for cloud-based TTS services, accounts, and credits, ensuring that all scripts and recordings never leave the device.

How it works

The application uses a native Swift runtime built on MLX to run Qwen3-TTS models locally. It avoids Python wrappers or local servers, instead utilizing a first-party engine that streams audio chunks to keep memory usage low and flat regardless of output length. It supports three primary modes: using built-in voices, designing a voice via plain-language descriptions, and cloning a voice from an audio reference.

Who it’s for

Content creators and developers using Apple Silicon Macs (macOS 26+) and iPhone 15 Pro or newer (iOS 26+) who require high-performance, private, on-device text-to-speech generation.

Highlights

  • Local-first Privacy: All generation, history, and voice data are stored locally; no data is sent to a cloud server.
  • Voice Design & Cloning: Ability to create voices from text descriptions or clone existing voices from audio files (WAV, MP3, AIFF, M4A, FLAC).
  • Multilingual Support: Automatic detection for ten languages, including English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish, and Italian.
  • High Performance: Optimized for Apple Silicon, achieving faster-than-realtime generation on an 8GB M2 Mac.
  • Long-form Projects: Handles scripts over 900 characters by streaming segments sequentially into a single finished file.
  • AI Watermarking: Includes an inaudible AudioSeal watermark and provenance notes in WAV files (version 3.0).

Related

  • Project
    leemysw/yovoice

    An open-source voice creation tool for macOS and Windows that turns text into natural speech locally, featuring voice cloning and emotion control without cloud API costs.

  • Project
    soniqo/speech-swift

    Speech Swift is an open‑source Swift package that provides on‑device speech AI for Apple Silicon. It bundles many modern ASR, TTS, translation, and voice‑agent models (e.g., Qwen3‑ASR, Whisper‑ASR, Omnilingual ASR, Qwen3‑TTS, CosyVoice, PersonaPlex) built on MLX and Core ML, so you can run inference locally with no cloud calls. Each model is a separate Swift‑Package‑Manager product, making it easy to add only what you need. The repo includes quick‑start code, SwiftUI UI helpers, Homebrew install, and extensive documentation.

  • Project
    Blaizzy/mlx-audio-swift

    A modular Swift SDK for audio processing on Apple Silicon, providing a unified API for TTS, STT, VAD, and audio codecs using the MLX framework.

  • Project
    Blaizzy/mlx-audio

    An audio processing library optimized for Apple Silicon that provides fast inference for text-to-speech, speech-to-text, speech-to-speech, and music generation.

  • Project
    james-see/ltx-video-mac

    A native SwiftUI macOS app for local AI video generation on Apple Silicon, supporting LTX and MiniMax H3 models with integrated audio and voiceover tools.