🔊 MOSSlanding

The best voice-cloning TTS on the planet. Runs on your machine. Costs nothing.

Clone any voice from a 5-second sample. Speak 31 languages. Zero cloud. Zero subscription. Zero compromise.

---

🚀 Quick Install

Debian / Ubuntu (APT)

# One command:
curl -fsSL https://xirtus.github.io/MOSSlanding/install-apt.sh | sudo bash
mosslanding

Or manually:

wget https://github.com/xirtus/MOSSlanding/releases/latest/download/mosslanding_latest_amd64.deb
sudo dpkg -i mosslanding_latest_amd64.deb
mosslanding

Arch Linux (Pacman)

# Option 1: Download from GitHub Releases
wget https://github.com/xirtus/MOSSlanding/releases/latest/download/mosslanding-1.0.0-1-x86_64.pkg.tar.zst
sudo pacman -U mosslanding-1.0.0-1-x86_64.pkg.tar.zst
mosslanding

# Option 2: Build from AUR (coming soon — PKGBUILD in packaging/arch/)
# git clone https://aur.archlinux.org/mosslanding.git
# cd mosslanding && makepkg -si

macOS

# Download DMG from Releases
open https://github.com/xirtus/MOSSlanding/releases/latest
# Drag Mosslanding.app to /Applications

Windows

# Download and run installer from Releases
Invoke-WebRequest -Uri "https://github.com/xirtus/MOSSlanding/releases/latest/download/Mosslanding-Setup.exe" -OutFile "Mosslanding-Setup.exe"
.\Mosslanding-Setup.exe

---

📦 Package Formats

---

🎯 Features

🎤 Voice Cloning — Drop in any 5–30 second audio clip and it instantly learns the voice. Accent, tone, cadence — all of it.

🌍 31 Languages — English, Chinese, Japanese, French, Spanish, Arabic, Hindi, Korean and 23 more. Switch mid-sentence.

🎯 Pinyin / IPA / Phoneme control — Nail every pronunciation. Perfect for names, technical terms, multilingual scripts.

⏸️ Pause control — [pause 1.5s] anywhere in your text. Precise timing.

⚡ GPU Accelerated — NVIDIA RTX 2070+ with bfloat16 precision. Apple Silicon via Metal MPS.

🔒 100% Private — Everything runs on your machine. Your voice samples never leave your device.

---

🖥️ Platforms

Requirements: 8GB+ VRAM, 16GB system RAM, Python 3.10+

---

🔧 Build from Source

git clone https://github.com/xirtus/MOSSlanding
cd MOSSlanding

# Linux
cd release/linux && ./install.sh && mosslanding

# Build packages
cd release/packaging && ./build_all.sh
# Output in release/dist/

---

📁 Repository Structure

release/
├── linux/          # Linux app source + install scripts
├── mac/            # macOS SwiftUI app + DMG builder
├── windows/        # Windows app + PowerShell installer
├── appimage/       # AppImage builder
├── packaging/      # Native package builders
│   ├── arch/       # PKGBUILD for Arch Linux / pacman
│   ├── debian/     # DEBIAN/ control files for .deb
│   ├── apt-repo/   # APT repository generator + one-liner installer
│   └── build_all.sh # Build everything in one shot
└── dist/           # Built packages output

---

🏗️ Built on MOSS-TTS

This app is a native desktop launcher for the MOSS-TTS family by OpenMOSS — state-of-the-art open-source TTS that outperforms closed-source models like Doubao and Gemini 2.5 Pro in subjective evaluations.

Models used:

• OpenMOSS-Team/MOSS-TTS-v1.5 (~4-6 GB) — voice cloning & direct generation
• OpenMOSS-Team/MOSS-VoiceGenerator (~1-2 GB) — voice design from text descriptions
• OpenMOSS-Team/MOSS-Audio-Tokenizer (~1 GB) — audio encoding/decoding

---

🚢 Creating a Release

1. Push a version tag:

   git tag v1.0.0
   git push origin v1.0.0
   

2. GitHub Actions automatically builds .deb and .pkg.tar.zst and attaches them to the release.

3. (Optional) Deploy APT repo to GitHub Pages:

   cd release/packaging/apt-repo
   ./setup-repo.sh 1.0.0
   git checkout gh-pages
   cp -r repo/* .
   git add . && git commit -m "apt repo v1.0.0" && git push
   

4. (Optional) Submit PKGBUILD to AUR for yay -S mosslanding support.

---