comfyui_maya1_tts_alun

🎙️ Maya1 TTS ComfyUI Custom Node

中文文档 | English

A high-quality text-to-speech (TTS) ComfyUI custom node powered by the Maya1 model, featuring multiple voice styles and flexible parameter configuration.

✨ Features

• 🎵 High-Quality Speech Synthesis: Based on the advanced Maya1 voice generation model
• 🎨 Multiple Voice Styles: 5 built-in voice presets (male, female, different ages)
• 🔧 Flexible Parameters: Adjustable temperature, chunk length, and more
• 💾 Automatic Model Management: Models auto-download and cache to ComfyUI models directory
• ⚡ Smart Caching: Models load once for improved generation efficiency
• 🌐 International Support: English UI and logging

📦 Installation

Method 1: Clone directly to ComfyUI custom nodes directory

cd ComfyUI/custom_nodes/
git clone https://github.com/ruanjianlun/comfyui_maya1_tts_alun.git
cd comfyui_alun_maya1
pip install -r requirements.txt

Method 2: Manual Installation

1. Copy the entire project folder to ComfyUI/custom_nodes/ directory
2. Install dependencies:

pip install -r requirements.txt

Dependencies

Main dependencies include:

• transformers - Hugging Face model library
• torch - PyTorch deep learning framework
• snac - SNAC audio codec
• soundfile - Audio file processing
• numpy - Numerical computing

🚀 Usage

1. Add Node in ComfyUI

1. Launch ComfyUI
2. Right-click on canvas, select Add Node -> audio -> maya1 -> Maya1 Text to Speech
3. The node will appear on the canvas

2. Configure Node Parameters

The node provides the following parameters:

Parameter	Type	Description	Default
text	String	Text to convert to speech	"Hello, this is a test..."
voice_preset	Dropdown	Voice style preset	Male-Mature
chunk_length	Integer	Text chunk length (characters)	50
temperature	Float	Generation temperature (0.1-1.0, higher = more random)	0.4
custom_description	String (Optional)	Custom voice description	Empty

3. Voice Style Presets

5 built-in voice style presets:

• Male-Mature: Male voice around 30 years old, American accent, normal pitch, warm timbre
• Female-Gentle: Female voice around 20 years old, American accent, slightly high pitch, soft timbre
• Male-Energetic: Male voice around 20 years old, American accent, high energy, bright timbre
• Female-Professional: Female voice around 30 years old, British accent, professional tone, clear timbre
• Neutral-Broadcast: Neutral voice, standard accent, clear articulation, balanced timbre

4. Custom Voice Description

If presets don't meet your needs, use the custom_description parameter for custom voice characteristics:

Realistic male voice in the 40s age with british accent. 
Deep pitch, authoritative tone, slow pacing.

Description template:

• Age: in the 20s/30s/40s age
• Accent: american/british/australian accent
• Pitch: high/normal/low/deep pitch
• Timbre: warm/bright/soft/clear timbre
• Pacing: fast/moderate/slow/conversational pacing
• Style: energetic/professional/gentle/authoritative tone

📁 Project Structure

comfyui_alun_maya1/
├── __init__.py              # Node registration entry file
├── maya_tts_node.py         # Main node implementation
├── config.py                # Configuration file (model paths, constants, etc.)
├── requirements.txt         # Python dependenc