In this episode of the ComfyUI course, you’ll learn how to install the Nunchaku custom node, understand int4 and fp4 Nunchaku models, and use ready-made Nunchaku workflows to significantly reduce VRAM usage and speed up image generation. This tutorial is ideal for users with RTX 30, 40, and 50 series GPUs, especially those working with limited VRAM.

By the end of this video, you will know:

• How to install the ComfyUI Nunchaku node using the easy installer
• Which Nunchaku model type (int4 vs fp4) is right for your GPU
• Where to place Nunchaku diffusion models, CLIP encoders, and VAEs
• How to load and use Z Image, Qwen Image, and Flux Nunchaku workflows
• How Nunchaku compares with other models in speed and quality
• How to convert an existing workflow into a Nunchaku workflow
• How to fix common Nunchaku and Flux issues

Workflows download links (Discord, GitHub Releases, Google Drive): https://github.com/pixaroma/pixaroma-workflows

Unlock exclusive perks by joining our channel:
https://www.youtube.com/channel/UCmMbwA-s3GZDKVzGZ-kPwaQ/join

Grab your Pixaroma merch from here: https://www.youtube.com/@pixaroma/store

#comfyui #comfyuitutorial #textoimageai
Chapters:
00:00 What Is Nunchaku in ComfyUI
00:41 Updating ComfyUI + Nunchaku System Requirements
02:24 Installing the Nunchaku Custom Node
04:15 Downloading Free Episode 02 Workflows
05:58 Nunchaku Z Image + int4 vs fp4 + Comparison
11:00 Nunchaku Qwen Image
16:03 Nunchaku Flux 1 Dev
20:00 Seed Variation
21:22 Fixing Flux Data Type Errors
21:53 Compare all 3 Models + Song :)