Qwen/Qwen-Image-2.1-PE-I2I

🤖 ModelScope | 🤗 HuggingFace | 📑 Blog | 🖥️ Demo | 🫨 Discord

Introduction

We are excited to open-source Qwen-Image-2.1, a unified text-to-image generation and image editing model in the Qwen family. With just 7B parameters in its visual generation component (32 Single-Stream DiT layers), Qwen-Image-2.1 balances generation quality, inference efficiency, and versatility.

Four key improvements define this release:

• Compact and Efficient — A lightweight architecture with mixed-granularity attention and prefix KV cache reuse delivers strong image quality at low computational cost.
• Native Transparency, Unified Creation and Editing — Generate regular or transparent (RGBA) images from text, edit transparent layers, and extract subjects from photographs—all in one model.
• Versatile Editing — Support up to 10 reference images, specify local edits via circles, painted annotations, or separate masks, and preserve identity for people and products.
• Realistic Textures and Refined Aesthetics — Improved typography, portrait lighting, and fine details for more visually compelling results.

Qwen-Image-2.1-PE-I2I

Image editing prompt rewriting model for Qwen-Image-2.1. A fine-tuned Qwen3.5-VL 9B that takes a vague editing instruction plus input image(s) and produces a precise, actionable prompt suitable for downstream image editing.

For more details, see the GitHub repo and Blog.

Quick Start

Installation

pip install transformers>=5.4.0 torch>=2.4.0 accelerate pillow

Usage with Transformers

import json
import torch
from PIL import Image
from transformers import AutoModelForImageTextToText, AutoProcessor

model_id = "Qwen/Qwen-Image-2.1-PE-I2I"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
model_id, dtype=torch.bfloat16, device_map="auto"
).eval()

# Load the system prompt shipped with the model
import huggingface_hub
sys_prompt_path = huggingface_hub.hf_hub_download(model_id, "system_prompt.txt")
system_prompt = open(sys_prompt_path).read().strip()

input_image = Image.open("input.png").convert("RGB")
user_prompt = "make the sky sunset"

messages = [
{"role": "system", "content": [{"type": "text", "text": system_prompt}]},
{"role": "user", "content": [
{"type": "image", "image": input_image},
{"type": "text", "text": user_prompt},
]},
]

inputs = processor.apply_chat_template(
messages, add_generation_prompt=True, tokenize=True,
return_dict=True, return_tensors="pt", enable_thinking=True,
).to(model.device)

with torch.no_grad():
out = model.generate(
**inputs, max_new_tokens=24000,
do_sample=True, temperature=1.0, top_p=0.95, top_k=20,
)
gen = processor.tokenizer.decode(
out[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True
)

# Split thinking from the answer
thinking, _, answer = gen.partition("")
result = json.loads(answer.strip())
print(result)
# {"rewritten_prompt": "...", "wh_ratio": "", "ratio_follow": ""}

Multi-Image Editing

The model supports multiple input images — referred to as `, `, etc.:

images = [Image.open("portrait.png").convert("RGB"),
Image.open("scene.png").convert("RGB")]

messages = [
{"role": "system", "content": [{"type": "text", "text": system_prompt}]},
{"role": "user", "content": [
{"type": "image", "image": images[0]},
{"type": "image", "image": images[1]},
{"type": "text", "text": "Place 's subject into 's scene"},
]},
]

Integration with Diffusers

import torch
from PIL import Image
from diffusers import QwenImage21Pipeline

# Assuming `result` and `input_image` from above
prompt = result["rewritten_prompt"]

pipe = QwenImage21Pipeline.from_pretrained(
"Qwen/Qwen-Image-2.1", torch_dtype=torch.bfloat16
).to("cuda")

image = pipe(
prompt=prompt,
image=input_image,
num_inference_steps=40,
generator=torch.Generator("cuda").manual_seed(42),
).images[0]

image.save("rewritten_edit.png")

Output Format

The model outputs a JSON object after a `` reasoning block:

{
"rewritten_prompt": "",
"wh_ratio": "",
"ratio_follow": ""
}

• rewritten_prompt — the expanded prompt to pass to the image editing model
• wh_ratio — aspect ratio chosen by the model (e.g. "16:9"), when the task creates a new composition
• ratio_follow — inherit aspect ratio from an input image (e.g. ""), when editing in-place

wh_ratio and ratio_follow are mutually exclusive — exactly one carries a value.

License

This model is licensed under the [Qwen Research License Agreement](./LICENSE).