16
Qwen-Image is here (aussie.zone)
submitted 1 week ago* (last edited 1 week ago) by Eyekaytee@aussie.zone to c/localllama@sh.itjust.works

๐Ÿš€ Meet Qwen-Image โ€” a 20B MMDiT model for next-gen text-to-image generation. Especially strong at creating stunning graphic posters with native text. Now open-source.

๐Ÿ” Key Highlights:

๐Ÿ”น SOTA text rendering โ€” rivals GPT-4o in English, best-in-class for Chinese

๐Ÿ”น In-pixel text generation โ€” no overlays, fully integrated

๐Ÿ”น Bilingual support, diverse fonts, complex layouts

๐ŸŽจ Also excels at general image generation โ€” from photorealistic to anime, impressionist to minimalist. A true creative powerhouse.

Blog: https://qwenlm.github.io/blog/qwen-image/

Hugging Face: https://huggingface.co/Qwen/Qwen-Image

Model Scope: https://modelscope.cn/models/Qwen/Qwen-Image/summary

GitHub: https://github.com/QwenLM/Qwen-Image

Technical Report: https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/Qwen_Image.pdf

WaveSpeed Demo: https://wavespeed.ai/models/wavespeed-ai/qwen-image/text-to-image

Demo: https://modelscope.cn/aigc/imageGeneration?tab=advanced

top 11 comments
sorted by: hot top controversial new old
[-] SW42@lemmy.world 7 points 1 week ago

But can it make a Poster of Tank man on Tianamen Square?

[-] Nomad@infosec.pub 6 points 1 week ago

In fact it can but it makes Chinese propaganda pics first. Just ask and it makes pictures of a soldier in uniform on a tank with a red star. But ask more specific about the massacre and this will fall out:

[-] SW42@lemmy.world 3 points 1 week ago

Huh :) the output quality is actually pretty impressive. It rivals Flux for sure.

[-] Eyekaytee@aussie.zone 3 points 1 week ago
[-] Nomad@infosec.pub 2 points 1 week ago

One of the demo links from above

[-] Eyekaytee@aussie.zone 0 points 1 week ago* (last edited 1 week ago)

TBH I haven't used any local image generators like Flux etc in a long time so I'm not even sure how to input this in, I think LM Studio is still a way off

What do you use?

[-] SW42@lemmy.world 3 points 1 week ago* (last edited 1 week ago)

I actually run Qwen locally using LM Studio. Even then it won't say anything that it deems as "controversial". If you want to use Flux, I'll share my LM Studio workflow later when I'm back at my ML Workstation

[-] Eyekaytee@aussie.zone 2 points 1 week ago

You can't run Qwen-image in LM Studio? It doesn't support image generation:

[-] SW42@lemmy.world 4 points 1 week ago* (last edited 1 week ago)

If you want to use Flux,

Oh, sorry - Brainfart. I meant Comfy UI... I do use Qwen on LM Studio and it's censored.

[-] Eyekaytee@aussie.zone 2 points 1 week ago

ah awesome, let me check! i used qwen for ages before flicking over to GLM and I've not been impacted but it's not like i ask about chinese government things very often

[-] SW42@lemmy.world 2 points 1 week ago

i finally got to the workstation. after instaling ComfyUI you need to add the ComfyUI-GGUF Node https://github.com/city96/ComfyUI-GGUF if you're using Apple Silicon - i didn't manage to get it to work otherwise because of the data type conversion.

Finally this is the Workflow I use for image generation: https://voidbin.com/paste/6f15026e-d18d-4542-97aa-2a93acc97af6 just save it as a json.

this post was submitted on 05 Aug 2025
16 points (70.0% liked)

LocalLLaMA

3525 readers
6 users here now

Welcome to LocalLLaMA! Here we discuss running and developing machine learning models at home. Lets explore cutting edge open source neural network technology together.

Get support from the community! Ask questions, share prompts, discuss benchmarks, get hyped at the latest and greatest model releases! Enjoy talking about our awesome hobby.

As ambassadors of the self-hosting machine learning community, we strive to support each other and share our enthusiasm in a positive constructive way.

Rules:

Rule 1 - No harassment or personal character attacks of community members. I.E no namecalling, no generalizing entire groups of people that make up our community, no baseless personal insults.

Rule 2 - No comparing artificial intelligence/machine learning models to cryptocurrency. I.E no comparing the usefulness of models to that of NFTs, no comparing the resource usage required to train a model is anything close to maintaining a blockchain/ mining for crypto, no implying its just a fad/bubble that will leave people with nothing of value when it burst.

Rule 3 - No comparing artificial intelligence/machine learning to simple text prediction algorithms. I.E statements such as "llms are basically just simple text predictions like what your phone keyboard autocorrect uses, and they're still using the same algorithms since <over 10 years ago>.

Rule 4 - No implying that models are devoid of purpose or potential for enriching peoples lives.

founded 2 years ago
MODERATORS