CG Dreams New Image Generator Beats EVERYTHING!

Damn this gives you a lot of control, and the 3D model feature is very interesting

Please replace anything dystopian with utopian, replace cyber- and steam punk with solarpunk etc, lets visualize and manifest an inspiring future, attention is the origin of creation ❤

NSFW is paywall, almost DOA, at least they allow it.

ok my jaw dropped when I saw the style transfer unto a normal image!

I was having a compostive dream. Lucid is what I mean. Apparently! Imagination is more powerful than knowledge. I can not suppress this hype anymore!!!

This comment was generated by Ai

It’s really hitting me now that games are gonna get less expensive to make and ppl are still gonna pay $70 for them.

Everyone keep your ear to the ground. Sam Altman tweeted that OpenAI would be rolling out the new voice features this week to paid subscribers!

I don’t know how the premium version is, but the free version is absolute rubbish. I uploaded a photo head and shoulders and I had bears shoulders and it told me that it was sensitive content and wouldn’t generate my image. Then three times I tried to select a photo from my photo library and it shows the wrong photo from my Library, not impressed

We need a Dall E-4 already.

While I do certainly appreciate broad coverage of the AI space, I'm not really sure that the hyperbolic description of this product is really justified by the content within. It's certainly high quality so far as text to image generation goes, but in my opinion it doesn't really offer anything that can't be done with existing solutions, and in particular highly custom open source workflows. I think a more useful piece of content would be to aggregate the general trends in ComfyUI workflows that you tend to see and a practical look at integrating existing tools like texturing software or Blender into open source workflows which offer much more control than anything hidden behind a website or API.

I struggle getting any AI to generate images that aren't "concept art". E.g. a mountain wall with certain lighting as background for a video game stage. Adding descriptions just reshuffles the scene. Hope such control becomes practical (simple) in near future, if it's not already.

What model does it use? Looking at the way it writes prompts, it could be some SD3, Kolors, or Auraflow variant. Or are they using their own proprietary diffusion model like Midjourney?

That looks like SD to me, and the filters are different LORAs.

You got a sub for me, had no idea this was out there good looks

Not bad considering it’s free. Midjourney really need to hurry up and release a new model

How does the queue worked? Signed up and paid, but I am still in the queue waiting for minutes for each image? Is this supposed to take so long?

Its unavailable in my country 😢

Does it generate text?

No, way anyone expects to get $19/mo value out of a member subscription. I’m very curious to see if anything come of it. 😅