How do I fix missing node errors in ChronoEdit?

Open ComfyUI Manager and click "Install Missing Custom Nodes." This will automatically detect and install any nodes the workflow requires. Restart ComfyUI after installation.

ChronoEdit ComfyUI Workflow: Easy Image Editing Guide (2026)

Q: What models does ChronoEdit require?

ChronoEdit requires five model types: a diffusion model (chrono_edit_14B_fp8_e4m3fn.safetensors or the GGUF version), a LoRA (chronoedit_distill_lora.safetensors), a text encoder (umt5_xxl_fp8_e4m3fn_scaled.safetensors), CLIP Vision (clip_vision_h.safetensors), and a VAE (wan_2.1_vae.safetensors).

Q: What settings should I use without the ChronoEdit LoRA?

If you disable the LoRA loader, set steps to 20, CFG to 3.0, and shift to 5.0 for best results without the distillation LoRA.

What ChronoEdit Can Do

ChronoEdit speciality: Physics-aware image editing, style transfer, and character-consistent transformations — all in ComfyUI with a ready-made workflow you can download below.

Physics-based editing: boy shoots a basketball — the ball goes in perfectly

Original image: a boy near a basketball hoop, before ChronoEdit processing — Original

OUTPUT

ChronoEdit output: boy shoots basketball into the hoop with realistic physics

Style transfer: portrait transformed into a trendy fashion look with red hair and sunglasses

Original portrait of a girl, before ChronoEdit fashion style transfer — Original

OUTPUT

ChronoEdit output: girl with red hair, stylish sunglasses, and fashion backdrop

Character consistency: same character transformed into a Japanese PVC figure style

Input: original portrait of a woman — Input

Output: same character rendered as a Japanese PVC scale figure — Japanese PVC scale figure

Introduction to ChronoEdit

ChronoEdit is an advanced image editing model designed to work with ComfyUI. It allows you to modify, enhance, and transform images using smart AI-driven transformation techniques — from physics-aware edits to full style transfers — while maintaining subject consistency across changes.

Unlike simple prompt-based editors, ChronoEdit understands spatial relationships and real-world physics, making it ideal for edits where objects need to move realistically or characters need to be transformed while preserving their identity.

Prerequisites: Prepare ComfyUI

Before starting, make sure ComfyUI is updated to the latest version to ensure compatibility with ChronoEdit's nodes.

Open ComfyUI Manager

Click the Manager button at the top of the ComfyUI interface.

Click "Update All"

This updates ComfyUI's core files and all installed custom nodes. Required for ChronoEdit to work smoothly.

ComfyUI Manager showing the Update All button highlighted to update all nodes and ComfyUI core — Click "Update All" in ComfyUI Manager before setting up ChronoEdit

Download Workflow

Download the ready-made ChronoEdit workflow JSON file and drag it into your ComfyUI workspace to load it instantly.

💡 Tip: Drag and drop the downloaded JSON file directly onto the ComfyUI canvas — no need to use the Load button.

Download Required ChronoEdit Models

All models below are required for ChronoEdit to work. Download each file and place it in the correct folder inside your ComfyUI installation — the table shows both the download link and the exact destination folder.

Type	File	Folder	Download
Diffusion Model (FP8)	`chrono_edit_14B_fp8_e4m3fn.safetensors`	`models/diffusion_models/`	⬇ Get
Diffusion Model (FP16)	`chrono_edit_14B_fp16.safetensors`	`models/diffusion_models/`	⬇ Get
Diffusion Model (GGUF)	`ChronoEdit-14B-Q8_0.gguf`	`models/unet/`	⬇ Get
LoRA	`chronoedit_distill_lora.safetensors`	`models/loras/`	⬇ Get
Text Encoder	`umt5_xxl_fp8_e4m3fn_scaled.safetensors`	`models/text_encoders/`	⬇ Get
CLIP Vision	`clip_vision_h.safetensors`	`models/clip_vision/`	⬇ Get
VAE	`wan_2.1_vae.safetensors`	`models/vae/`	⬇ Get

Folder Structure

Your ComfyUI models directory should look like this after placing all files:

ComfyUI/
└── models/
    ├── diffusion_models/
    │   └── chrono_edit_14B_fp8_e4m3fn.safetensors
    ├── unet/
    │   └── ChronoEdit-14B-Q8_0.gguf
    ├── loras/
    │   └── chronoedit_distill_lora.safetensors
    ├── text_encoders/
    │   └── umt5_xxl_fp8_e4m3fn_scaled.safetensors
    ├── clip_vision/
    │   └── clip_vision_h.safetensors
    └── vae/
        └── wan_2.1_vae.safetensors

💡 Tip: You only need one diffusion model variant — choose FP8 for lower VRAM usage, FP16 for maximum quality, or GGUF for CPU-offload setups.

Step-by-Step: ChronoEdit ComfyUI Workflow Setup

Update ComfyUI

Open ComfyUI Manager → click Update All to update all nodes and core files.

Load the workflow

Drag and drop the ChronoEdit JSON file into the ComfyUI workspace to load it.

Install missing nodes

If red nodes appear, go to Manager → "Install Missing Custom Nodes" and install everything listed. Restart ComfyUI after.

Select models in each node

In the workflow, choose the correct model file in each loader node — diffusion model, CLIP Vision, VAE, LoRA, and text encoder.

Load your image

Use the Load Image node to import the image you want to edit.

Write your prompt

Add your edit instructions in the prompt node. ChronoEdit works best with focused, single-task prompts.

(Optional) Disable SageAttention

If you have not installed SageAttention or encounter errors, disable the patch sage attention node — the workflow still works without it.

Click "Queue Prompt" to run

ComfyUI will process your image through the ChronoEdit pipeline and output the result.

⚠️ Note: If you disable the LoRA loader (chronoedit_distill_lora), set steps: 20, CFG: 3.0, and shift: 5.0 for best results without the distillation LoRA.

ComfyUI workspace showing the ChronoEdit workflow loaded with all nodes connected and model files selected — ChronoEdit workflow loaded in ComfyUI with all nodes configured

Example Run: Edit an Image with ChronoEdit

Here is a complete example. Load the input image in the Load Image node, use the prompt below, and run the workflow.

Example Prompt — PVC Figure Style Transfer

Use the same character from the input image. Create a photorealistic Japanese anime PVC figure of an elegant woman in a flowing red kimono with golden floral embroidery and a gold obi. She has long black hair with golden ornaments and a confident gaze, surrounded by drifting cherry blossom petals. Displayed on a simple round base on a computer desk with blurred keyboard and monitor glow. Studio lighting, shallow depth of field, detailed painted PVC textures, realistic shadows, and a premium collectible look.

Result: same character transformed into a photorealistic Japanese PVC scale figure

Input image: original portrait of a woman used as ChronoEdit reference — Input Image

ChronoEdit output: woman rendered as a Japanese PVC scale figure with kimono and golden ornaments — Output — PVC Figure

More Example Prompts and Results

Sketch → Anime Fight Scene

Prompt

Convert the sketch into a black-and-white Japanese anime fight scene. Show two warriors mid-battle with intense poses, motion lines, and dramatic manga-style inking. Use bold lines, strong contrast, and expressive shading with cross-hatching and ink splashes. Emphasize action, emotion, and energy — flying hair, torn clothes, and glowing eyes. The scene should look like a cinematic manga panel full of impact and tension.

Input: pencil sketch of two figures for ChronoEdit anime conversion — Sketch Input

ChronoEdit output: sketch converted to black-and-white Japanese anime fight scene with manga inking — Anime Battle Scene

Physics-Based Edit: Boy Shoots Basketball

Prompt

A young boy throws a basketball toward the hoop, and the ball goes in perfectly.

Original image: boy near a basketball hoop before ChronoEdit animation — Original

OUTPUT

ChronoEdit output: boy shoots basketball — ball goes in the hoop with realistic trajectory

Fashion Style Transfer: Red Hair & Sunglasses

Prompt

Add stylish sunglasses to the girl and change her hair color to a vibrant, fashionable red. Replace the background with a stylish fashion-themed setting — like a modern studio or runway backdrop with soft lighting and elegant tones. Keep her facial features and pose the same, with glossy natural red hair and well-fitted, modern sunglasses.

OUTPUT

ChronoEdit output: same girl with vibrant red hair, stylish sunglasses, and a modern fashion backdrop

Tips and Troubleshooting

Missing node errors (red nodes in workflow)

Fix: Go to ComfyUI Manager → "Install Missing Custom Nodes". Install everything listed, then restart ComfyUI. All nodes should turn green.

Model not loading or not appearing in dropdown

Fix: Double-check file names and folder locations in the models/ directory. Refresh your browser after placing new files so ComfyUI picks them up.

SageAttention errors

Fix: Disable the patch sage attention node in the workflow. ChronoEdit will still function without it — SageAttention only speeds things up if it is correctly installed.

Poor quality results or unexpected outputs

Fix: Focus each prompt on a single main change. Complex multi-step prompts reduce quality. If not using the LoRA, set steps: 20, CFG: 3.0, and shift: 5.0.

One-task prompts work best. Here is a good example — changing only the camera angle of a running girl:

Focused single-task prompt

Change the running girl to a full side view, showing her running from left to right. Keep the same outfit, lighting, and background style, just adjust the camera angle to capture her in full profile as she runs.

Input: girl running toward the camera, original angle — Original angle

OUTPUT

Side profile output

Frequently Asked Questions

ChronoEdit is an advanced image editing model for ComfyUI that allows users to modify, enhance, and transform images using smart AI-driven techniques. It excels at physics-aware edits, style transfers, and character-consistent transformations — generating realistic outputs that understand spatial relationships.

ChronoEdit needs seven model files: a diffusion model (FP8, FP16, or GGUF), a LoRA (chronoedit_distill_lora.safetensors), a text encoder (umt5_xxl_fp8_e4m3fn_scaled.safetensors), CLIP Vision (clip_vision_h.safetensors), and a VAE (wan_2.1_vae.safetensors). Each goes in its matching ComfyUI/models subfolder.

Open ComfyUI Manager and click "Install Missing Custom Nodes." This automatically detects and installs any nodes the workflow requires. Restart ComfyUI after installation — all red nodes should turn green.

If you disable the LoRA loader (chronoedit_distill_lora), set steps to 20, CFG to 3.0, and shift to 5.0 for best results without the distillation LoRA. With the LoRA enabled, default settings work well.

Yes. ChronoEdit can transform static images into short video clips while maintaining realistic physics — such as a basketball going into a hoop, a character changing pose, or style transformations animated over frames.