Repository navigation
Make Flux Kontext workflow also work with qwen image edit 2509 ? #32
Description
The recent qwen edit 2509 is very good and is working in comfyui, needs the same inputs (image + text)
Is it possible to make a new tool that support it? or add it to the current kontext tool and rename it to Image editing tool
Activity
Haervwe commentedon Sep 27, 2025
Hi Chris!! The room should work if you change the workflow, but I recognize that is not clear in the docu, i'm. Working with one of the contributors in updating this room to support a more varied workflows selection! I'll be sure to add a qwen Image one! I couldn't find why the responses do not get to the UI on mobile on some I stances I didn't forgot lol 😂
iChristGit commentedon Sep 28, 2025
Hey! Thanks for the response, Il wait as much as needed, It would be helpful to change it on the fly to qwen.
Also, maybe consider creating a VibeVoice tool? It has amazing abilities and can do any language, a good system prompt can make sure the output is
[1]: <speaker 1 text>
[2]: <speaker 2 text>
and user can choose the mp3 for reference beforehand in comfyui, maybe worth adding?
@Haervwe
Haervwe commentedon Oct 10, 2025
Hi @iChristGit sorry ofr the delay i didn forget, was just bussy at work, i just finished some modifications to the Kontext pipe, and now focusing on gettting it to use qwen image, i think i should be straig forward, yes, i need some tts tools, the problem is that i cannot resolve having a built it way to play the files, but ywah its on the works!
Haervwe commentedon Nov 2, 2025
Hi ! @iChristGit ! Good news! I added qwen edit as a tool and vibe voice tts! Check them out and tell. Me what you think :)
iChristGit commentedon Nov 3, 2025
Hey!
Where can I find the VibeVoice TTS? does it auto generate the lyrics with appropriate tags and all? amazing!!
Haervwe commentedon Nov 3, 2025
Hey Chris! Here it is https://github.com/Haervwe/open-webui-tools/blob/main/tools/comfyui_vibevoice_tts_tool.py
It has two wf one for single and one for multi speaker, and es it instruct the LLM on how to use it :)