🎧 VoxForge Pro

Premium AI-Powered Audiobook Generator with Kokoro-82M & Chatterbox TTS

Transform your text and PDF documents into professional audiobooks with 47 premium AI voices across 6 languages. Fully offline after installation.

✨ Features

🎤 47 Premium Voices

• American English: 20 voices (11 female, 9 male)
• British English: 8 voices (4 female, 4 male)
• Brazilian Portuguese: 4 voices
• Chinese Mandarin: 8 voices
• Japanese: 5 voices
• Spanish: 3 voices

🤖 Dual TTS Engines

• Kokoro-82M: Fast, high-quality synthesis
• Chatterbox: Voice cloning with 5-10s reference audio

📄 Smart PDF Processing

• Digital PDFs: Direct text extraction
• Scanned PDFs: OCR with 15+ languages
• Automatic chapter detection
• Custom page range selection

🎚️ Professional Audio Control

• Adjustable speed (0.5x - 2.0x)
• Multiple output formats (MP3, WAV, FLAC, OGG)
• Configurable bitrate (64-320 kbps)
• Chapter-by-chapter generation

📚 Library Management

• Organized audiobook library
• Playback and download
• Usage statistics
• Duplicate detection

🚀 Quick Start

Via Pinokio (Recommended)

1. Install Pinokio
2. Search for "VoxForge Pro"
3. Click Install (downloads ~5GB)
4. Click Start
5. Open http://127.0.0.1:7860

Manual Installation

git clone https://github.com/Shinshekai/VoxForge-Pro.git
cd VoxForge-Pro

# Create virtual environment
python -m venv app/env
source app/env/bin/activate  # Linux/Mac
# or: app\env\Scripts\activate  # Windows

# Install dependencies
uv pip install -r app/requirements.txt

# Install PyTorch with CUDA
uv pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

# Run
python app/main.py

🎙️ Voice Catalog

🇺🇸 American English

Female	Male
Alloy, Aoede, Bella, Heart, Jessica	Adam, Echo, Eric, Fenrir, Liam
Kore, Nicole, Nova, River, Sarah, Sky	Michael, Onyx, Puck, Santa

🇬🇧 British English

Female	Male
Alice, Emma, Isabella, Lily	Daniel, Fable, George, Lewis

🇧🇷 Brazilian Portuguese

Female	Male
Camila, Dora	Alex, Santa

🇨🇳 Chinese Mandarin

Female	Male
Xiaobei, Xiaoni, Xiaoxiao, Xiaoyi	Yunjian, Yunxi, Yunxia, Yunyang

🇯🇵 Japanese

Female	Male
Alpha, Gongitsune, Nezumi, Tebukuro	Kumo

🇪🇸 Spanish

Female	Male
Dora	Alex, Santa

💡 Usage Tips

Multi-Voice Narration

Use speaker tags for different characters:

[Narrator]The door creaked open slowly.[/Narrator]
[Alice]Who's there? she whispered.[/Alice]
[Bob]It's just me, don't worry.[/Bob]

Best Practices

1. For clarity: Use af_heart or bf_emma
2. For drama: Use am_fenrir or bm_george
3. For children's books: Use af_kore or jf_nezumi
4. For audiobooks: Use am_michael or bf_alice

🔧 System Requirements

• OS: Windows 10/11, macOS, Linux
• RAM: 8GB minimum, 16GB recommended
• GPU: NVIDIA with 4GB+ VRAM (optional but recommended)
• Storage: 10GB for installation + space for audiobooks

🔌 API Reference

VoxForge Pro includes a Gradio API for programmatic access.

Python

from gradio_client import Client

client = Client("http://127.0.0.1:7860")
result = client.predict(
    text="Hello, this is a test.",
    voice="af_heart",
    speed=1.0,
    api_name="/synthesize"
)
print(result)  # Path to generated audio file

JavaScript

import { Client } from "@gradio/client";

const client = await Client.connect("http://127.0.0.1:7860");
const result = await client.predict("/synthesize", {
    text: "Hello, this is a test.",
    voice: "af_heart",
    speed: 1.0
});
console.log(result.data);  // Path to generated audio file

cURL

curl -X POST http://127.0.0.1:7860/api/synthesize \
  -H "Content-Type: application/json" \
  -d '{"data": ["Hello, this is a test.", "af_heart", 1.0]}'

📄 License

VoxForge Pro NON-COMMERCIAL EVALUATION LICENSE 1.0

• ✅ Free for personal and non-commercial use
• ❌ Commercial use of the software requires a separate license
• ❌ Commercial use of generated audio requires a separate license
• ❌ Voice cloning without consent is prohibited

See LICENSE for full details.

🙏 Credits

• Kokoro by hexgrad
• Chatterbox by Resemble AI
• PaddleOCR by PaddlePaddle
• Gradio by HuggingFace

---

Made with ❤️ by Shinshekai for storytellers everywhere