Kandinsky-5.0-I2I-Lite-sft-Diffusers

<div align="center"> <picture> <img src="assets/KANDINSKY_LOGO_1_BLACK.png"> </picture> </div> <div align="center"> <a href="https://habr.com/ru/companies/sberbank/articles/951800/">Habr</a> | <a href="https://ai-forever.github.io/Kandinsky-5/">Project Page</a> | <a href="https://github.com/kandinskylab/kandinsky-5/blob/main/paper.pdf">Technical Report</a> | <a href="https://github.com/ai-forever/Kandinsky-5">Original GitHub</a> | <a href="https://huggingface.co/collections/kandinskylab/kandinsky-50-image-lite-diffusers">🤗 Diffusers</a> </div>

---

<h1>Kandinsky 5.0 I2I Lite SFT – Diffusers</h1>

Kandinsky 5.0 is a family of diffusion models for video and image generation.

Kandinsky 5.0 Image Lite is a lightweight text-to-image (I2I) generation model with 6B parameters.

The model introduces several key innovations:

• Latent diffusion pipeline with Flow Matching for improved training stability
• Diffusion Transformer (DiT) as the main generative backbone with cross-attention to text embeddings
• Dual text encoding using Qwen2.5-VL and CLIP for comprehensive text understanding
• Flux VAE for efficient image encoding and decoding

The original codebase can be found at kandinskylab/Kandinsky-5.

Available Models

Kandinsky 5.0 Image Lite:

Examples

<table border="0" style="width: 90%; text-align: center; margin-top: 20px;"> <tr> <td style="width:45%; vertical-align:top;"> <img src="assets/generation_examples/images/7.jpg" style="width:48%; display:inline-block; vertical-align:top;"> <img src="assets/generation_examples/images/7_edit.png" style="width:48%; display:inline-block; vertical-align:top;"> <div style="margin-top:8px;"> Change the image style to realistic, draw the horse in the same pose, with a black mane and a giant snowball on its nose </div> </td> <td style="width:10%;"></td> <td style="width:45%; vertical-align:top;"> <img src="assets/generation_examples/images/6.jpg" style="width:48%; display:inline-block; vertical-align:top;"> <img src="assets/generation_examples/images/6_edit.png" style="width:48%; display:inline-block; vertical-align:top;"> <div style="margin-top:8px;"> Draw this elephant in the same hat inside a green tractor, driving across a yellow field. </div> </td> </tr> </table> <table border="0" style="width: 90%; text-align: center; margin-top: 20px;"> <tr> <td style="width:45%; vertical-align:top;"> <img src="assets/generation_examples/images/4.jpg" style="width:48%; display:inline-block; vertical-align:top;"> <img src="assets/generation_examples/images/4_edit.png" style="width:48%; display:inline-block; vertical-align:top;"> <div style="margin-top:8px;"> Поменяй чебурашку на крокодила Гену из мультфильма, позу и выражения лица оставь прежними </div> </td> <td style="width:10%;"></td> <td style="width:45%; vertical-align:top;"> <img src="assets/generation_examples/images/5.jpg" style="width:48%; display:inline-block; vertical-align:top;"> <img src="assets/generation_examples/images/5_edit.png" style="width:48%; display:inline-block; vertical-align:top;"> <div style="margin-top:8px;"> Change the image style to an oil painting; don't change the pose of the animals, but add pronounced oil painting elements - paint strokes </div> </td> </tr> </table>

Kandinsky5I2IPipeline Usage Example

import torch
from diffusers import Kandinsky5I2IPipeline
from diffusers.utils import load_image 
# Load the pipeline
model_id = "kandinskylab/Kandinsky-5.0-I2I-Lite-sft-Diffusers"
pipe = Kandinsky5I2IPipeline.from_pretrained(model_id)

_ = pipe.to(device='cuda',dtype=torch.bfloat16)
pipe.enable_model_cpu_offload()                                               # <--- Enable CPU offloading for single GPU inference

# Edit the input image
image = load_image(
    "https://huggingface.co/kandinsky-community/kandinsky-3/resolve/main/assets/title.jpg?download=true"
)

prompt = "Change the background from a winter night scene to a bright summer day. Place the character on a sandy beach with clear blue sky, soft sunlight, and gentle waves in the distance. Replace the winter clothing with a light short-sleeved T-shirt (in soft pastel colors) and casual shorts. Ensure the character’s fur reflects warm daylight instead of cold winter tones. Add small beach details such as seashells, footprints in the sand, and a few scattered beach toys nearby. Keep the oranges in the scene, but place them naturally on the sand."
negative_prompt = ""

output = pipe(
    image=image,
    prompt=prompt,
    negative_prompt=negative_prompt,
    guidance_scale=3.5,
).image[0]

Results

<table style="width:100%; text-align:center; margin-top:20px;"> <tr> <td> <img src="assets/sbs_image.png" width="100%"> </td> <td> <img src="assets/sbs_edit.png" width="100%"> </td> </tr> <tr> <td style="font-size: 1.1em; font-weight: 500; padding-top: 6px;"> Side-by-side evaluation of T2I SFT on PartiPrompts with extended prompts </td> <td style="font-size: 1.1em; font-weight: 500; padding-top: 6px;"> Side-by-side evaluation of I2I SFT on the Flux Kontext benchmark with extended prompts </td> </tr> </table>

Citation

@misc{kandinsky2025,
    author = {Alexander Belykh and Alexander Varlamov and Alexey Letunovskiy and Anastasia Aliaskina and Anastasia Maltseva and Anastasiia Kargapoltseva and Andrey Shutkin and Anna Averchenkova and Anna Dmitrienko and Bulat Akhmatov and Denis Dimitrov and Denis Koposov and Denis Parkhomenko and Dmitrii and Ilya Vasiliev and Ivan Kirillov and Julia Agafonova and Kirill Chernyshev and Kormilitsyn Semen and Lev Novitskiy and Maria Kovaleva and Mikhail Mamaev and Mikhailov and Nikita Kiselev and Nikita Osterov and Nikolai Gerasimenko and Nikolai Vaulin and Olga Kim and Olga Vdovchenko and Polina Gavrilova and Polina Mikhailova and Tatiana Nikulina and Viacheslav Vasilev and Vladimir Arkhipkin and Vladimir Korviakov and Vladimir Polovnikov and Yury Kolabushin},
    title = {Kandinsky 5.0: A family of diffusion models for Video & Image generation},
    howpublished = {\url{https://github.com/kandinskylab/Kandinsky-5}},
    year = 2025
}