[Feature Request] Support qwen-image-edit

[jsqh]

Required prerequisites

• [x] I have searched the Issue Tracker and Discussions that this hasn't already been reported. (+1 or comment there if it has.)
• [x] Consider asking first in a Discussion.

Motivation

在做图像相关的usecase，请问能否把阿里的qwen-image-edit模型加入到model中以便直接调用，感谢~

add qwen-image-edit as one toolkit, refer: https://qwen.ai/blog?id=7a90090115ee193ce6a7f619522771dd9696dd93&from=research.latest-advancements-list

Solution

No response

Alternatives

No response

Additional context

No response

[Wendong-Fan]

thanks @jsqh for opening this issue, I think qwen-image-edit is more suitable for integration as a toolkit rather than a language model, as it's specifically designed for image editing tasks and not for chat.

I've added this feature request to the current development sprint.

[JINO-ROHIT]

hey guys, i was playing around with the qwen pipeline for image edits for a bit, but im not sure if its a good idea to support this model for now.(maybe after it has an api)

reason being the model is extremely large, requires massive VRAM and almost nobody would be using the toolkit via diffusers locally imo, wdyt?

[Wendong-Fan]

Thanks for the feedback @JINO-ROHIT . If the model is currently too heavy and lacks an easy-to-use API, let's postpone integration.

@jsqh would it be possible to use this model as an isolated tool for now to satisfy your use case? Alternatively, are there other image editing models we should explore?