AI Generated Images
Community for AI image generation. Any models are allowed. Creativity is valuable! It is recommended to post the model used for reference, but not a rule.
No explicit violence, gore, or nudity.
This is not a NSFW community although exceptions are sometimes made. Any NSFW posts must be marked as NSFW and may be removed at any moderator's discretion. Any suggestive imagery may be removed at any time.
Refer to https://lemmynsfw.com/ for any NSFW imagery.
No misconduct: Harassment, Abuse or assault, Bullying, Illegal activity, Discrimination, Racism, Trolling, Bigotry.
AI Generated Videos are allowed under the same rules. Photosensitivity warning required for any flashing videos.
To embed images type:
“”
Follow all sh.itjust.works rules.
Community Challenge Past Entries
Related communities:
- [email protected]
Useful general AI discussion - [email protected]
Photo-realistic AI images - [email protected] Stable Diffusion Art
- [email protected] Stable Diffusion Anime Art
- [email protected] AI art generated through bots
- [email protected]
NSFW weird and surreal images - [email protected]
NSFW AI generated porn
Like others have said: If you have a decent GPU in your PC, automatic1111 with stable diffusion is absolutely worth a try. For windows, it's not that hard to install. Just follow the guide on github step by step.
With my old nvidia gtx 970, it took about 45 seconds to render an image. Anything newer will only grant you faster generation.
Do note that if you have an AMD gpu, it's a pain on windows currently. My current 6700XT is about as fast as my old 970 on windows.
On Linux, AMD works great (6 seconds for an image for me right now), but that requires some tinkering.
Look into fooocus it's great, I've been having a blast with all the different preset art styles that you can use and it's totally free.
Downside is you'll need a pretty beefy PC to generate images fast. I have a i7, 32GB RAM and 1060 GPU and it takes around 30 minutes to generate 2 highly detailed images.
Thanks for putting some numbers down. At what resolution does it take 30 mins?
mage.space is not bad, but they have been gradually getting worse over time. I recommend having your own Automatic1111 workbench if you can.
The enshittification was inevitable
that is exactly what a bot would say
If you have a big GPU stable diffusion webui/easy diffusion will run on your own PC. There is also an option to use CPU if you have time
https://tinybots.net/artbot and by extension https://aihorde.net hundreds of sd models, all Loras in civitai, completely free and without ads and free software.
Haha, how is that possible, free and no ads?! Incredible.
Spread the word. We have no ad budget since it's free 😁
holy carp.
how can be that good wut
must explore
ty for link
edit: just looked at that tinybots.net and looks amazing. pretend i didnt suggest anything lol
doubleedit: specifically because of the noobfriendly clause i actually will link mine: https://perchance.org/beautiful-people and no i get no moneys from perchance. free signup to not see ads and then its free infinite ai gen tho it's like a few years old aiwise i think. anyway, i think i made that one pretty nicely and very easy type prompt click button nice picture. i say this after having tried tinybots and it is less noobfriendly and did not allow me to scroll thru the pics i generated on computer. about to try it on phone but was enough i felt it wasn't 100% the absolute best no matter what and that mentioning anything else actually was feasible, lol. perchance is fun and free and infinite, but im going to be basically exploring that tinybots.net all night now lol. tho that not scrolling thing was weird. hopefully it doesn't do that on phone. also did not like how it ate my prompt after i generated it. also no easy way to download output without it being weird format and lame title was also offputting. first impression is it seems the most powerful and definite best, XCEPT for that it should be coded better and it's sad looking at it from BeautifulPeople having taken care of really basic parts like those. like the devs did a really good job on the back end and empowering it but the front is nonfun.
There's settings to retain your prompt. Also it's being reworked right now as well
Though I haven't yet really been able to get in to it, I love the complexity of your concept. Also I notice that in credits for AI Horde it says the person for Development and Maintenance is none other than db0. For me I would say that using my own thing where i have already done the steps of going thru the complexity and refining it to a feelgood interface makes it hard to jump in to something i have done 0 refinement to. So that is a challenge for me. I like though that you are making a system open to developers so even someone like me could, over time, make something based on your system that fulfills my needs. And I like your creative design concept of Kudos instead of Money; essentially valuing 'work' in a decentralized fashion. Basically, someone could make an app using aiHorde's ai gen, have each user that is creating images also using their data to build everyone's images, and ai horde is then a selfsustaining system and, if users are coding their apps correctly, there could even, for free, be more generative power than the power required by those using it; which would then make it feel better for 'everyone using it'. It's a very smart model and bypasses straight spending of cash to buy stuff to feed a system. But looking in to the 'workers' aspect, it feels like it isn't quite there and, instead, requires big computers with big graphics cards to do 'the work', and those with smaller instead are 'manually rewarded for tasks' which wouldn't work on something like an infinitely scaling Kotlin App for Android. Is there yet no way to make a comfortable Kotlin App for Android that generates images while feeding back in more work than it takes while also not heating up people's phones and melting them? If possible, high five you best. tho it seems not there and that the workers aspect has too hefty requirements and then relegated to big people with rooms of multiple high end computers empowering things instead of true decentralized power. am i correct?
Will you be able to integrate SD3.0 seamlessly and stay up to date?
cool thing tho. gj making it
wow i just could not get in to tinybots. it seems so powerful yet all my art came out like steamy dungpile. did work better on phone but was just so much easier to make better pics on my own thing. also been learning SDXL on https://playground.com/create which is free too
Night Cafe. Free to use but limited uses. Can log in daily for 5 coins, do other tasks for some more. I found it pretty easy to save up a few days then go on sprees of trying things. There is a totally free option there too but it doesn't have some of the finesse of the coin use ones.
There are also daily challenges and community challenges, which I think are great to help with ideas. And some of the communities if come across are very helpful with things to add, or add as negative prompts, etc. There is a "group" of artists who leave their prompts open to see too. So when you look through or vote on the other art, you can click to see info and those folks have the prompts they used visible for that piece, in case it's helpful to you.
Bing, just need to create a login
Specifically Copilot, I believe
I will call the bingalator my co pilot only after I am cold and dead.