KokoroMac - GitLite

KO | EN

gitlite — search

#chatgpt #computer-vision #android #machine-learning #mysql #typescript #video #api #proxy #reactjs

KokoroMac

(★ 14)

Offline Voice Studio for Your Mac.

파일 탐색기

ZIP 다운로드

- - Contents.json
- - AppSettings.swift
  - AppVersion.swift
  - PythonServerCode.swift
  - SetupManager.swift
  - Theme.swift
  - TTSManager.swift
- - VoiceModel.swift
- - AudioPlayerView.swift
  - LeftPanel.swift
  - MacTextEditor.swift
  - MainView.swift
  - RightPanel.swift
  - SetupWizardView.swift
  - VoicePickerPopover.swift
- KokoroMacApp.swift
- - contents.xcworkspacedata
- project.pbxproj
- KokoroMaclogo.jpg
.gitignore
LICENSE.txt
README-ZH.md
README.md

# CDN으로 사용하기

jsDelivr

jsDelivr는 공개 GitHub 리포지토리를 별도 설정 없이 CDN으로 즉시 서빙합니다. 버전과 파일을 고르면 웹페이지에 바로 붙일 수 있는 링크와 예시 코드가 만들어집니다.

버전

파일

링크

예시

명령어 용어집

이 문서에서 사용된 명령어를 모아봤습니다. 낯선 명령어가 있다면 펼쳐서 확인해보세요.

🔍

`espeak-ng`

설명 보기 ▼

다국어를 지원하는 소프트웨어 음성 합성기.
관련 항목: `speak-ng`, `espeak`.
더 많은 정보: <https://github.com/espeak-ng/espeak-ng/blob/master/src/espeak-ng.1.ronn>.

espeak-ng "{{text}}"

텍스트를 음성으로 출력:

echo "{{text}}" | espeak-ng

`stdin`의 텍스트를 음성으로 출력:

espeak-ng -f {{path/to/file}}

파일([f]ile) 내용을 음성으로 출력:

🔍

`git checkout`

설명 보기 ▼

Checkout a branch or paths to the working tree.

git checkout -b {{branch_name}}

Create and switch to a new branch:

git checkout -b {{branch_name}} {{reference}}

Create and switch to a new branch based on a specific reference (branch, remote/branch, tag are examples of valid references):

git checkout {{branch_name}}

Switch to an existing local branch:

🔍

`xattr`

설명 보기 ▼

확장 파일 시스템 속성을 다루는 유틸리티.
더 많은 정보: <https://keith.github.io/xcode-man-pages/xattr.1.html>.

xattr -l {{file}}

주어진 파일의 key:value 확장 속성 나열:

xattr -w {{attribute_key}} {{attribute_value}} {{file}}

주어진 파일에 속성 작성:

xattr -d {{com.apple.quarantine}} {{file}}

주어진 파일에서 속성 삭제:

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<p align="center">
  <img src="public/KokoroMaclogo.jpg" alt="KokoroMac Logo" width="64" />
  <br />
  <h1 align="center">KokoroMac</h1>
  <p align="center">Local, Open-Weight Text-to-Speech Studio for Mac.</p>
  <p align="center">
    <a href="https://github.com/arinltte/KokoroMac/releases/latest"><img src="https://img.shields.io/github/v/release/arinltte/KokoroMac?style=flat-square&color=blue" alt="Latest Release" /></a>
    <a href="https://github.com/arinltte/KokoroMac/blob/main/LICENSE"><img src="https://img.shields.io/github/license/arinltte/KokoroMac?style=flat-square&color=green" alt="License" /></a>
    <img src="https://img.shields.io/badge/macOS-14.0%2B-blue?style=flat-square" alt="macOS" />
    <img src="https://img.shields.io/badge/100%25-Offline-brightgreen?style=flat-square" alt="Offline" />
  </p>
</p>

<p align="center">
  <a href="./README.md">English</a> | <a href="./README-ZH.md">中文文档</a>
</p>

## Introduction

KokoroMac is a native macOS desktop application that brings the power of [Kokoro TTS](https://github.com/hexgrad/kokoro) directly to your Apple Silicon Mac. It provides a beautiful, easy-to-use interface for generating high-quality speech from text completely offline.

Powered by the state-of-the-art, open-weight 82M parameter Kokoro model, KokoroMac requires no internet connection, no API keys, and no data leaves your machine after the initial setup. Just type, generate, and export.

<video src="https://github.com/user-attachments/assets/edb29c04-dcc5-4044-a78b-90a58a0b9e3a" controls width="800"></video>

## ✨ Top Features

*   🚀 **100% Offline & Private:** Runs entirely on-device. Your text and generated audio never leave your Mac.
*   🧠 **Powered by Kokoro-82M:** Leverages the lightweight, open-weight TTS model for natural-sounding, high-fidelity voices.
*   ⏸️ **Mathematically Exact Pauses:** Visual UI blocks for inserting precise pauses (Short, Medium, Long). The backend trims AI decay and splices exact digital silence for perfect audiobook pacing.
*   🗣️ **Phoneme Overrides:** Force exact pronunciations for tricky words or proper nouns using the International Phonetic Alphabet (IPA) without breaking text formatting.
*   🎨 **Beautiful Native UI:** Features a "True Dark" mode, ambient themes, and an interactive waveform player with seek and hover timestamps.
*   ⚡ **Smart Resource Management:** Automatically unloads AI models from RAM after idle periods to keep your Mac running smoothly.
*   📦 **One-Click Setup & Silent Updates:** Automatically provisions Python environments, installs dependencies, and downloads AI models. Upgrading to new versions automatically migrates backend scripts without requiring Terminal commands.

## ⚙️ Requirements

KokoroMac is designed to be self-sufficient and will automatically install Python, `espeak-ng`, and all AI models during the initial setup. However, the following are required beforehand:

*   **macOS 14.0 (Sonoma)** or later.
*   **Apple Silicon (M1/M2/M3/M4)** Mac.
*   **[Homebrew](https://brew.sh/):** The app uses Homebrew to install Python 3.11 and `espeak-ng`. (If missing, the setup wizard will provide the exact terminal command to install it).
*   **Internet Connection:** Only required during the very first launch to download dependencies and model weights.

## 📥 Installation

1.  Download the latest `KokoroMac.dmg` from the [Releases page](https://github.com/arinltte/KokoroMac/releases/latest).
2.  Drag `KokoroMac.app` to your Applications folder.
3.  **Bypass macOS Gatekeeper:** Because the app is not notarized, macOS will block it on first launch. Open your Terminal and run:
    ```bash
    xattr -rd com.apple.quarantine /Applications/KokoroMac.app
    ```
4.  Launch the app and follow the on-screen Setup Wizard.

## 🚀 Getting Started

1.  **Type or Paste Text:** Enter your script into the Speech Text editor.
2.  **Add Directives:** Use the Speech Toolbox to insert visual Pause blocks (0.5s, 1.0s, 2.0s) or apply Phoneme Overrides for tricky pronunciations.
3.  **Select a Voice:** Choose from a variety of US and UK English voices.
4.  **Adjust Speed:** Fine-tune the speech rate using the Speed slider (0.5x to 1.5x).
5.  **Generate:** Click "Generate Speech" and listen to the output in the Audio Preview waveform player.
6.  **Export:** Save your generated audio as a high-quality `.wav` file.

## 🔒 Data & Privacy

KokoroMac respects your privacy. All processing happens locally. No telemetry, no cloud APIs.

| Location | Contents |
| :--- | :--- |
| `~/.KokoroMac/Python` | Isolated Python 3.11 virtual environment and pip packages. |
| `~/.KokoroMac/Cache` | HuggingFace Hub cache containing the Kokoro-82M model weights, voices, and spaCy NLP models. |
| `~/.KokoroMac/Temp` | Temporary `.wav` files generated during playback (auto-cleaned on app launch). |
| `~/.KokoroMac/server.py` | The local FastAPI backend script that bridges the Swift UI with the Python TTS engine. |

## Uninstallation

To completely remove KokoroMac and all downloaded AI models/dependencies, run:
```bash
rm -rf /Applications/KokoroMac.app
rm -rf ~/.KokoroMac
```

## 🤝 Contributing

Contributions are welcome. Whether it's a bug report, a feature suggestion, a documentation improvement, or a pull request — all are appreciated.

To contribute:
1.  Fork the repository.
2.  Create a feature branch: `git checkout -b feature/your-feature-name`
3.  Commit your changes with a clear message.
4.  Open a pull request against `main` with a description of what you changed and why.

To report a bug or request a feature, open an [issue](https://github.com/arinltte/KokoroMac/issues). Please include your macOS version and steps to reproduce for bug reports.

## 📄 License & Acknowledgements

The KokoroMac application source code is licensed under the [MIT License](./LICENSE).

*   **Core TTS Engine:** This application is a graphical wrapper for the incredible [Kokoro TTS](https://github.com/hexgrad/kokoro) model and library, created by `@hexgrad` and contributors. The Kokoro model weights, Python library, and underlying audio data are licensed under the Apache License 2.0.
*   **Audio Splicing & Pause Pipeline:** The architectural approach for mathematical pause timing, audio trimming, and backend splicing integrated in v0.4.0 was heavily inspired by the community approaches documented in [nazdridoy/kokoro-tts](https://github.com/nazdridoy/kokoro-tts).

We deeply appreciate the open-source AI community for making local inference possible.