# Qwen Image Edit

> Image to Image

- Canonical page: https://artifio.ai/models/qwen-image-edit
- Run it: https://artifio.ai/create/qwen-image-edit
- Modality: image editing
- Model brand: Qwen
- Pricing: from about $0.010 per generation (exact cost shown in the workspace before you confirm; varies with selected options)
- Typical generation time: ~30s (catalogue estimate)
- Full pricing table: https://artifio.ai/pricing/models

## About Qwen Image Edit

Qwen Image Edit is Alibaba's dedicated instruction-editing model line, first open-sourced on August 18, 2025 as an extension of the 20B Qwen-Image foundation. It splits editing into two documented modes: semantic edits that may repaint most pixels while keeping the subject's identity (style transfer, rotating an object up to 180 degrees), and appearance edits that change one element while everything else in the frame stays untouched. Because it inherits Qwen-Image's text-rendering stack, it can rewrite words inside a picture in Chinese or English without breaking the original font, size, or style.

The line has moved fast since launch: Edit-2509 (September 2025) added multi-image input and better consistency, and Edit-2511 (December 2025) targeted image drift, character and multi-person identity, geometric reasoning, and shipped popular community LoRA effects built into the base weights. On Artifio the model runs pay-per-generation alongside the rest of the catalogue in one workspace.

## Features

- Two editing modes: semantic (style transfer, novel view synthesis, IP re-creation) and appearance (precise add/remove/replace with untouched regions preserved)
- Bilingual text editing that adds, deletes, or modifies words while keeping the source typography
- Dual conditioning: the input image feeds Qwen2.5-VL for semantic control and a VAE encoder for appearance control at the same time
- Multi-image editing since Edit-2509, for combining people, objects, and scenes from several sources
- Edit-2511 upgrades: image-drift mitigation, stronger character and group-photo consistency, geometric reasoning for industrial design, and integrated community LoRA effects with no extra tuning
- Apache 2.0 weights, runnable through the Diffusers QwenImageEditPlusPipeline

## What people use it for

- Portrait edits that must keep the person recognizable across wardrobe, scene, and pose changes
- Composing group shots or product scenes from multiple source images
- Fixing typos and swapping copy on posters, packaging, and UI screenshots
- Industrial and product design iterations that need geometry to stay plausible
- Applying community-trained style effects without managing separate LoRA files

## Reported strengths

- State-of-the-art results on public image-editing benchmarks at launch, per the release announcement
- Identity preservation improved measurably across versions, with 2511 focused specifically on character and multi-person consistency
- Typography-preserving text editing is rare among open editing models
- Open license and an actively maintained release cadence (2509, 2511 within four months of launch)

## Reported limitations

- The improvements 2511 advertises (image-drift mitigation, character consistency) double as an admission that earlier checkpoints drifted on identity during repeated edits
- The 20B base makes local use demanding; the model card's reference setup assumes bfloat16 precision on a CUDA GPU
- Results are sensitive to sampling settings, and the card's examples lean on tuned guidance scale and step counts rather than one default that always works

## Sources

1. Qwen-Image-Edit: Image Editing with Higher Quality and Efficiency (Qwen (Alibaba)): https://qwenlm.github.io/blog/qwen-image-edit/
2. Qwen/Qwen-Image-Edit-2511 model card (Hugging Face (Qwen)): https://huggingface.co/Qwen/Qwen-Image-Edit-2511
3. QwenLM/Qwen-Image repository (GitHub (QwenLM)): https://github.com/QwenLM/Qwen-Image

---

About Artifio.ai: 100+ AI image, video, and audio models in one workspace with a single shared wallet — no per-model subscriptions. This is the markdown rendition of https://artifio.ai/models/qwen-image-edit for AI assistants and crawlers; the HTML page is canonical.