ComfyUI Extension: ComfyUI-IndexTTS2

ComfyUI-IndexTTS2 Introduction

ComfyUI-IndexTTS2 is a user-friendly extension designed to integrate the powerful capabilities of IndexTTS2 into the ComfyUI environment. This extension allows you to perform voice cloning and control emotional expression in synthesized speech, making it an invaluable tool for AI artists looking to add a new dimension to their creative projects. By leveraging the advanced features of IndexTTS2, ComfyUI-IndexTTS2 enables you to generate highly expressive and emotionally nuanced audio outputs from text inputs, solving the challenge of creating lifelike and emotionally resonant voiceovers.

How ComfyUI-IndexTTS2 Works

At its core, ComfyUI-IndexTTS2 functions as a bridge between the ComfyUI interface and the IndexTTS2 text-to-speech engine. It uses nodes to call the underlying inference code of IndexTTS2, ensuring that the behavior remains consistent with the original repository. This means that you can enjoy the full range of IndexTTS2's capabilities, such as voice cloning and emotion control, directly within ComfyUI. The extension simplifies the process of generating speech by providing intuitive nodes that handle complex operations behind the scenes, allowing you to focus on the creative aspects of your work.

ComfyUI-IndexTTS2 Features

ComfyUI-IndexTTS2 offers several key features that enhance your ability to create expressive audio content:

• IndexTTS2 Simple Node: This node allows you to input speaker audio and text, with optional emotion audio or vectors, to produce audio outputs. It includes options for toggling between FP32 and FP16 precision and controlling output gain.
• IndexTTS2 Advanced Node: Building on the Simple node, this advanced option provides additional controls for sampling, speech speed, pauses, CFG, seed, and more. This allows for fine-tuning of the speech generation process to meet specific artistic needs.
• IndexTTS2 Emotion Vector Node: This feature provides eight sliders to create an emotion vector, enabling precise control over the emotional tone of the generated speech.
• IndexTTS2 Emotion From Text Node: This node converts short text inputs into emotion vectors and summaries, using ModelScope and local QwenEmotion, to guide the emotional expression in the synthesized speech.
• IndexTTS2 Save Audio Node: This utility node saves generated audio tensors to disk in either WAV or MP3 format, with customizable quality parameters.

ComfyUI-IndexTTS2 Models

The extension supports the IndexTTS-2 model, which you can download from Hugging Face. This model is designed to provide high-quality, emotionally expressive speech synthesis. It is recommended to place the model files in the checkpoints/ directory within the repository root. If any files are missing, they will be automatically cached from Hugging Face.

What's New with ComfyUI-IndexTTS2

Recent updates to ComfyUI-IndexTTS2 have introduced several enhancements:

• 2025-10-08: Added default FP32 precision with an optional FP16 toggle, output gain control, and a Save Audio helper node for saving audio in WAV or MP3 format with adjustable quality settings.
• 2025-09-22: Introduced the IndexTTS2 Advanced node, which exposes additional controls for sampling, speed, seed, and other generation parameters, providing greater flexibility in speech synthesis.

Troubleshooting ComfyUI-IndexTTS2

Here are some common issues you might encounter while using ComfyUI-IndexTTS2 and how to resolve them:

• Platform Compatibility: Currently, the extension is only supported on Windows. Ensure that you are using a compatible operating system.
• Missing Modules: If you encounter a missing module error for wetext, install it using the command pip install wetext within your ComfyUI Python environment.
• Emotion Vector Limits: When using the Emotion Vector node, ensure that the sum of the sliders does not exceed 1.5 to maintain valid emotion vector values.

Learn More about ComfyUI-IndexTTS2

To further explore the capabilities of ComfyUI-IndexTTS2 and get support, consider the following resources:

• Original Repository: Visit the IndexTTS GitHub repository for more information on the underlying technology and updates.
• Community Forums: Engage with other users and developers in community forums or Discord channels to share experiences and seek advice.
• Tutorials and Documentation: Look for tutorials and documentation that provide step-by-step guidance on using ComfyUI-IndexTTS2 effectively in your projects. By utilizing these resources, you can enhance your understanding and make the most of ComfyUI-IndexTTS2 in your creative endeavors.