ComfyUI: Configure InfiniteTalk for talking-head generation

MDMZ·Dec 2025

Demo summary

The user demonstrates setting up the InfiniteTalk model loader, uploading a source image and audio file, and calculating the required frame count for a 25-second clip.

Step-by-step

Select a model in the InfiniteTalk model loader
Select the corresponding version in the One Video model loader
Import your AI avatar image using the Load Image node
Upload your audio file into the Load Audio node
Calculate the required frames by multiplying audio duration by 25
Enter the result into the Max Frames node

Options

Choose the Q8 model for faster generation and lower compute requirements
Choose the FP8 model for higher quality output
Select between 480p or 720p versions when using the FP8 model

Watch out for

You must match the model versions; if you use the Q8 InfiniteTalk model, you must use the Q8 One Video model

Tips

Use the Q8 version if you have limited computing power
Note that InfiniteTalk generates at 25 frames per second when calculating your frame count

Highlights

“Technically, this workflow can generate videos of infinite length.”

All demos from “Create Realistic AI AVATARS That Look And Talk EXACTLY Like You! | FREE ComfyUI Tutorial”

1:121:10Generate AI avatar images with Nano Banana ProThe creator demonstrates how to use Nano Banana Pro on Higgsfield.ai to generate a realistic portrait by uploading a selfie, entering a prompt, and selecting 2K or 4K quality settings.Higgsfield· AI Avatar Generator
3:520:45Create AI avatars using Quen in ComfyUIA walkthrough of using the Quen image edit workflow template inside ComfyUI to generate a free AI avatar image from a source selfie and text prompt.ComfyUI· AI Avatar Generator
5:421:22Configure InfiniteTalk for talking-head generationCurrentThe user demonstrates setting up the InfiniteTalk model loader, uploading a source image and audio file, and calculating the required frame count for a 25-second clip.ComfyUI· AI Avatar Video Generator
7:321:10Animate AI avatar with InfiniteTalk settingsThe creator adjusts the audio scale for expressiveness, adds a descriptive prompt for movement style, and configures the One Video Sampler steps before running the final generation.ComfyUI· AI Avatar Video Generator
Watch “Create Realistic AI AVATARS That Look And Talk EXACTLY Like You! | FREE ComfyUI Tutorial” →

AI Avatar Video Generator

25:063:04Create talking avatars with InfiniteTalkThe demonstration shows how to sync a static character image with an audio file to create a talking avatar using the InfiniteTalk model and Wan 2.1.pixaroma
5:421:22Configure InfiniteTalk for talking-head generationCurrentThe user demonstrates setting up the InfiniteTalk model loader, uploading a source image and audio file, and calculating the required frame count for a 25-second clip.MDMZ
7:321:10Animate AI avatar with InfiniteTalk settingsThe creator adjusts the audio scale for expressiveness, adds a descriptive prompt for movement style, and configures the One Video Sampler steps before running the final generation.MDMZ
0:441:10Generate a talking head video with Infinite Talk in ComfyUIThe creator demonstrates the basic Infinite Talk workflow by uploading a portrait image and an audio file, adjusting the prompt and resolution, and rendering a lip-synced video.pixaroma
7:081:19Automating podcast videos with Ollama and Chatterbox SRTA workflow demonstration showing how to use Ollama for text generation and Chatterbox SRT for text-to-speech to automatically drive the InfiniteTalk avatar length and audio sync.Benji’s AI Playground
8:431:16Generating a 52-second avatar videoThe user walks through the step-by-step process of inputting a script, selecting a reference voice for cloning, and initiating the sampling process to generate a nearly one-minute video.Benji’s AI Playground
7:580:38Generating the final talking videoThe creator shows the KSampler settings using Lightning LoRA for fast 7-step generation and displays the final 53-second talking head result.Aiconomist