Official 2025 Dall-E Mega Gallery 🍀

[taniajorgeytico]

You create images that represent me
I might be a little crazy, but in a good way.
And the brain one perfectly represents me when I watch K-dramas and something surprising happens hahaha.

[taniajorgeytico]

I was trying to make it more natural and less “magazine-like,” but the result is terrible. They still look artificial but with unhealthy symptoms

[mitchell_d00]

“Less polished natural woman in a leather cap”

Real bad eye errors…

[taniajorgeytico]

I don’t know what’s going on… DALL-E has a virus that’s spreading to the images hahaha, they all look like they need to be admitted to the hospital.

[Tina_ChiKa]

With all due respect, we are here in the gallery with AI Art - don’t we also have a monthly theme?

We have many excellent topics on DALL-E generation issues that should not be overlooked!

[Daller]

Collection of Dall-E 3 prompting tips, issues and bugs

There were some updates recently, but still generation errors and poor quality. Template effect, artifacts, color abarations etc. I get sometimes black and white pictures. And the template effects get stronger, entire faces and heads of photo models show up on monster bodies. Or completely chaotic images. Nothing in the prompts should trigger this. Still nonsense text show up in some images. Some images have a confetti kind of grain and noise. And the linguistic system witch should fill obj…

[Tina_ChiKa]

@PTXeno, for example, has made a good suggestion and just because someone is “quieter”, the contribution should not be overlooked.

[taniajorgeytico]

Ohh! Is there a topic?
I didn’t even remember! I got caught up learning and trying to improve. I’ll get back to the topic now, sorry.

[PTXeno]

Thank you very much for your consideration; I don’t take it for granted, and I truly appreciate it, because, after all, we are strangers. Interacting on the internet is always tough and complicated. We are not designed for this kind of interaction, and humans adapt very slowly. If my post got lost among the others, it’s also by choice: I care deeply about things going as well as possible for everyone, and I’m aware that there’s a real person on the other side. I just want every opportunity to connect to not be wasted. And, tying back to the topic of adaptation, I prefer to take it slow and let myself fade into the crowd of posts if I don’t feel confident enough to engage. PS: Obviously, I’ve made plenty of mistakes on the internet in the past.

[Tina_ChiKa]

Oh, well, you’re welcome!

I agree with you and I understand you very well. I’m actually reluctant too but, admittedly, I really like your pictures and your ideas shouldn’t go down.

Things “should go well” for you too

[mitchell_d00]

No the topic “ monthly theme? ” is not a thing anymore, it is just a general gallery now.
@Tina_ChiKa & @taniajorgeytico

Other topics will have this tag “gallery “

https://community.openai.com/tag/gallery

[crunchybomb07]

“Bringing ideas to life through a seamless blend of creativity and technology. The holographic workstation and modular design reflect the future of innovation—personalized, adaptive, and inspiring. A vision in progress.”

[Tina_ChiKa]

Thank you @mitchell_d00 , I really wasn’t sure anymore

@taniajorgeytico
I didn’t mean why you felt addressed, sorry for the misunderstanding
Indeed, I didn’t mean anyone in particular

My observation was only:
I just noticed the group dynamic where there was a bit too much focus on image generation errors at times

• Well, I just wanted to point out that there are topics from members at forum who deal with this intensively and that is where their expertise can be found.
  We shouldn’t overlook this, their time and work goes into it!

And I thought it would be useful to discuss these errors there

What a gallery means to me:

I draw myself and for me a gallery is a place where ideas can be exchanged and creativity can flow!
Where we can inspire each other and present our own works - even if it’s not perfect!

Keyword - artistic freedom

[Tina_ChiKa]

Well, bear with me - one of the things I’m working on in my research is interaction dynamics in AI-human interactions.
That is why I am somewhat sensitive to the flow and patterns in dynamic systems

[hugebelts]

Edit: This time this was done from scratch in only one iteration using a MetaPrompt.

[PTXeno]

[hugebelts]

[lordelfa]

I told Chat what this was for and told it to impress. This was its pick and this was its prompt:

“A stunning, surreal landscape that fuses nature and technology, showcasing the evolution of AI and its harmonious integration with the natural world. The scene features towering trees with branches made of intricate, glowing circuitry, and leaves that shimmer with digital patterns. In the sky, majestic clouds morph into swirling data streams, and a futuristic cityscape rises in the distance with buildings that blend organic shapes and advanced technology. A serene river flows through the scene, its waters reflecting holographic fish and luminous flora along the banks. The atmosphere is otherworldly yet balanced, with a warm, golden light suggesting a future where nature and technology coexist in harmony. The image embodies the awe-inspiring progress of AI, symbolized by the intricate details and seamless fusion of organic and mechanical elements.”

[wincupgo]

My personal opinion. In the next half a year OpenAI will have to go back to a more progressive version of DALL-E and reduce the censorship level to a minimum. The reason will be DeepSeek’s Janus-Pro image generation model.
Let’s look into the gaming industry. In Asia, there is no large-scale agenda against the sexualisation of characters as there is in the West. But these games are actively sold in our market.
What’s to stop them from replicating this in the image generation industry?

And another test of the old and current version:

[igochan]

I don’t think Dalle’s problem is censorship.
Dalle3 is lagging compared to other models. It is not only this filter that prevents making more realistic images, the training of the images conditions the result a lot. Have you ever tried to make an image of a river in the jungle? In three out of four images, I get a hut despite asking for only jungle. Or the overhead lights in some images. I asked for an image of someone in the dungeons of a castle and dalle puts a skylight in the ceiling for more light.
OpenAI has clearly focused a lot on chatgpt, but Flux and SD3.5 are leaving Dalle3 far behind.