Skip to content

README.md

Latest commit

 

History

History
103 lines (72 loc) · 3.86 KB

README.md

File metadata and controls

103 lines (72 loc) · 3.86 KB

Kandinsky 5 Video for ComfyUI

Kandinsky 5 ComfyUI Workflow

Description

This project provides a workflow for generating videos and images using the Kandinsky 5 model within the ComfyUI environment.

Updates

  • 🔥 2025/11/28: Image 2 Image and Text 2 Image pipelines added.
  • 🔥 2025/11/13: Image 2 Video pipeline added.

Installation and Setup

1. Install ComfyUI

If you don't have ComfyUI installed yet, follow these steps:

# Clone ComfyUI repository
git clone https://github.com/comfyanonymous/ComfyUI.git
cd ComfyUI

# Install dependencies
pip install -r requirements.txt

# Launch ComfyUI
python main.py

2. Clone this repository into the ComfyUI custom_nodes folder:

# Navigate to ComfyUI custom_nodes folder
cd ComfyUI/custom_nodes

# Clone this repository and install requirements for model
git clone https://github.com/kandinskylab/kandinsky-5.git kandinsky
cd kandinsky
pip install -r requirements.txt

3. Load the Workflow

Launch ComfyUI (typically available at http://127.0.0.1:8188)

In the ComfyUI interface, click the "Load" button

Select the appropriate JSON file for your task from kandinsky-5/comfyui directory:

Workflow File Mode Description
kandinsky5_lite_T2V.json Text-to-Video Generate a video sequence from a text prompt
kandinsky5_lite_I2V.json Image-to-Video Animate a starting image using a text prompt
kandinsky5_lite_T2I.json Text-to-Image Generate a static image from a text prompt
kandinsky5_lite_I2I.json Image-to-Image Edit or transform an image using text instructions

The workflow will load into the ComfyUI interface

4. Download Checkpoints

Download the required models and place them in the appropriate folders.

  1. Run download_models.py It will download models and encoders to ./weights directory.
  2. Rearrange them to comfyui paths(text_encoders/diffusion_models/vae).
./weights/text_encoder -> ComfyUI/models/text_encoders/text_encoder
./weights/text_encoder2 -> ComfyUI/models/text_encoders/text_encoder2
./weights/vae -> ComfyUI/models/vae/hunyuan_vae
./weights/flux/vae -> ComfyUI/models/vae/flux_vae

ComfyUI/
├── models/
│   ├── text_encoders/          # For text_encoder and text_encoder2 models
│   ├── diffusion_models/       # For kandinsky5lite_*2v*_*.safetensors models  
│   └── vae/
│        ├── hunyuan_vae         # For vae model
│        └── flux_vae           # For flux/vae model

5. Configure Parameters

After loading the workflow, configure the following parameters:

Main Parameters

Parameter Description Recommended Value
Prompt Text description for video generation Your descriptive text
Negative Prompt What to exclude from generation Unwanted elements description
Width/Height/Length Output video size 768x512x121 for 5s or 768x512x241 for 10s, Width and Height should be divisisble by 128 for 10s model. In case of image 2 video aspect ratio of result is close to start image ratio and area is close to 768*512
Steps Number of generation steps 50, 16 for distilled version
CFG Scale 1.0 for distilled16steps and nocfg, 5.0 for sft and pretrain models
Scheduler Scale Noise scheduler scale 5.0 for 5s, 10.0 for 10s and image 2 video

6. Notions

Image2video checkpoint can work in text2video mode too.

In text2video mode image could be used as input for extend_prompt node. In this case generated video follow style of input image.