Which Python version does Kokoro Speaker support?

Kokoro Speaker does not currently support Python 3.13 or above inside ComfyUI. Use an earlier Python version (3.10 or 3.11 recommended) for stable operation.

How to Install Kokoro Speaker in ComfyUI (2026 Guide)

Introduction

Quick answer: Kokoro Speaker is a free, open-source TTS model that runs entirely on your machine inside ComfyUI. No API key, no cloud costs, full privacy — just natural-sounding speech from any text you type.

Looking to add natural-sounding text-to-speech (TTS) to your projects? Kokoro Speaker is a powerful local TTS tool that works perfectly with ComfyUI. This guide explains what Kokoro Speaker is and gives you clear steps to install and use it, including how to generate unique blended voices.

What Is Kokoro Speaker?

Kokoro Speaker is a small, fast, and open-source text-to-speech model. It can turn any written text into natural-sounding speech directly on your computer. Because it runs locally, you have full control and complete privacy — no data is sent to any external server.

You can also use Kokoro Speaker to create different types of voices or generate long audio clips with ease. It supports multiple voice styles and integrates directly into your existing ComfyUI workflows.

Kokoro Speaker node overview inside a ComfyUI workflow showing text input and audio output — Kokoro Speaker running inside a ComfyUI workflow

How to Install Kokoro Speaker on ComfyUI

Follow these steps to install Kokoro Speaker with ComfyUI. The whole process takes about 5 minutes.

💡 Tip: Download the workflow first — you can drag and drop it directly into ComfyUI after installation.

Step 1 — Install ComfyUI

First, make sure ComfyUI is installed on your system. Use the latest version for the best compatibility. Read the full ComfyUI installation guide →

Step 2 — Open Custom Nodes Manager

Open ComfyUI. Go to the Manager and click on "Custom Nodes Manager".

Step 3 — Search and Install Kokoro

In the search bar, type:

Kokoro

Find and install "comfyui-kokoro" (developed by Stav Sapir).

Step 4 — Restart ComfyUI

After installation completes, restart ComfyUI to finish the setup and register the new nodes.

Step 5 — Load the Kokoro Speaker Workflow

Download the basic workflow below and drag-and-drop it into ComfyUI, or use the Load button inside the interface.

Download Kokoro Workflow

Step 6 — Input Your Text

Enter the text you want to convert to speech in the Kokoro Speaker node's text input field.

Kokoro Speaker ComfyUI node showing the text input field where you type the words to be converted to speech — Enter your text in the Kokoro Speaker node's input field

Step 7 — Run the Workflow

Click "Queue Prompt" to run. On first use, Kokoro will automatically download the required model weights — this may take a few minutes depending on your connection speed.

⚠️ Warning: The first run may take several minutes as Kokoro downloads its model files. Subsequent runs are fast.

Creating Unique Voices with Kokoro Speaker

Kokoro Speaker allows you to combine different voices for creative results. To blend two voices together:

Open ComfyUI and search for the "Kokoro Speaker Combiner" node.
Add two instances of the Kokoro Speaker node and connect each to the Combiner.
Run the workflow to produce a unique blended voice output.

ComfyUI workflow showing two Kokoro Speaker nodes connected to a Kokoro Speaker Combiner node to blend two voices — Connect two Kokoro Speaker nodes to the Combiner to create a unique blended voice

Four sample workflows and audio examples are available below to demonstrate unique voice combinations. Download any workflow and drag it into ComfyUI to try it immediately.

Voice Samples

Example 1 — Female Voice A

Download Workflow

Example 2 — Male Voice A

Download Workflow

Example 3 — Female Voice B

Download Workflow

Example 4 — Male Voice B

Download Workflow

Important Notes

⚠️ Warning: Kokoro Speaker does not currently support Python 3.13 or above inside ComfyUI. Use Python 3.10 or 3.11 for smooth, stable operation.

If you run into installation issues, check the ComfyUI Troubleshooting Guide or verify that the custom node was installed correctly via the Custom Nodes Manager.

Frequently Asked Questions

Kokoro Speaker is a lightweight, open-source text-to-speech model that runs entirely locally inside ComfyUI. It converts written text to natural-sounding speech without sending data to any external server.

Open the ComfyUI Custom Nodes Manager, search for "Kokoro", install "comfyui-kokoro" by Stav Sapir, then restart ComfyUI. Load the provided workflow and enter your text to generate speech.

Yes. Use the Kokoro Speaker Combiner node in ComfyUI. Connect two Kokoro Speaker nodes to the Combiner and run the workflow to blend two voices into a unique output.

Kokoro Speaker does not currently support Python 3.13 or above inside ComfyUI. Use Python 3.10 or 3.11 for stable operation.

Yes. Kokoro Speaker is fully open-source and free. It runs locally on your machine with no subscription or API cost.

Conclusion

Kokoro Speaker brings advanced and customizable local TTS capabilities to ComfyUI users. By following these steps, anyone can install Kokoro Speaker, generate natural speech from text, and even create entirely new blended voices.

Make the most of local TTS and unlock new creative possibilities in voice generation — completely free, completely private, all on your own machine.