Regards for all your love and support ❤
avoiding negatives can improve result on SDXL
Description
somehow i messed up CLIPs and stuff, choose yourself what you like more, but probably its overtrained and LastFame and CrystalClear are the best Models out there anyways :D
FAQ
Comments (18)
Bro was holding out PEAK on us for like a whole day! What the hell? I'd say I want my money back, but I know damn well I ain't ever contributed nothing lmao
Could you tell us what was changed/fixed in the Pure fix version?, I did not see any mention in description so asking here.
Thanks again.
It says in the About This Version section on the right.
@skullzy77 thanks!
After testing: Pure_Fix version is actually very good and most importantly FUN to use! 💜
Yes, Crystal Clear was so good already, that kind of hard to top it.. but i feel this has lots of potential for those who enjoy trying complex prompts and this latest versions delivers lots of variety and surprises (in a good way), so it's for those who have tried many checkpoints and are always hungry for new ones to try. ^^
Understands my silly prompts well (locally i am using BREAK syntax node in comfyUI because my prompts are so long often hih), and makes fun scenarios with variety of results, exactly what i want as txt2img addict who can just render images for hours . =)
Visually great and anatomy is done quite nicely, dont even need hand fixes most times, eyes are always looking great and lively. Easy in general to get "life" in the images so not look fake.
Great job and a nice addition to the epiCRealismXL family from one of the big pioneers and heroes of the community. Thanks!
* humble courtsy bow *
Thanks for taking the time to review this; I really appreciate it.
@epinikion Was my pleasure. ^^ I was just supposed to test it a bit to see how it compared to CC, but it was so fun to use that ended up making images all night until morning, one of them sleepless nights well spent.. because it's one of those models that keep giving interesting results. Simple prompts work as expected, but it's very good with complex ones, blending in concepts and emotions very well. Not that many models do that equally well, there are a few, each with their unique great qualities. Yours has also the trademark look of your models (obviously :P)..
Also, to anyone crazy enough reading my long rants: in this model no need for a long list of "masterpiece, ultra-detailed, etc etc.." in the prompts, can just go straight to business and quality is superb in 4K images if your workflow and settings are done properly. So no need to add also Embeddings that has all that spam. This model gives good results with just the prompt; subject, situation/scenario, lighting. That's it. Simple and comfy.
Reflections and subsurface scatterings etc look SO SHARP and clean. Very cinematic and cool results without having to be expert, so i'd say this is a good model even for beginners, any new one doing AI art will feel like an artist with this one, because it makes good quality without having to sweat. Yes, a good stack of Loras help, but i did try without my usual PowerLoraLoader gang (10 Loras that are like best friends, always with me lmao), and results were nice, there is some good stuff baked and trained in clearly, your experience shows itself. So it works very well "out of the box" even for new people with relatively simple workflow. Ease of use, easy to get in and do cool images.
tldr: So it's easy to have fun Model for beginners, but also it has that ceiling to do complex prompts for experienced and curious people.. finds that it can do many things that not many can do as well, or most other models has to like swap a million settings and Loras, but this makes things easy. I'd actually recommend this for a friend who asks a cool model to try as their 1st one. They would get addicted on making images also all day like me haha.
..you know, i often end up fast going DMD2 based workflow, i have been enjoying using this in normal traditional settings, even right now batch of images are rendering.. i can't stop, plz help! ^^
Have winter fun holidays times @epinikion & friends, cool lurkers and other people of this community! 🩷❄️
Sorry for the late review, I was busy testing out a ton of prompts with this version.
And wow. I'm just blown away by how artistic it is and how well it follows my ideas.
Just wow... new version is really amazing
prompt adherence feels pretty much the same as Crystal clear
I'm getting the same face with every generation. LastFame is still the best!
Hola, Alguien puede decirme como descargar epicrealismxl y llevarlo a focus tarda mucho en cargarse e foocus.
puedes explicarme como le hicieron
me parece que depende de la estabilidad del la conexion de internet, hay que bajar el archivo y subrlo en focus pero no cambiar de pestaña de navegador
Just came by after using this model and wanted to leave a comment letting you know this is fantastic and I enjoy the model very much. Output quickly, produces great results, handles Loras very well, overall it is just a lot of fun with excellent results. Aldo handles different resolutions very well. Good job thanks.
Thank you for the kind words. I’m pleased if my models are useful to someone.
cool cool cool
Details
Files
epicrealismXL_pureFix.safetensors
Mirrors
epicrealismXL_pureFix.safetensors
epiCRealism_XL_KiSS.safetensors
epiCRealism_XL_Pure_fix.safetensors
epiCRealism_XL_V8.safetensors
epiCRealism_XL_V8_KiSS.safetensors
epicrealism_xl_pure_fix.safetensors
epiCRealism_XL_Pure.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_pureFix.safetensors
epiCRealism_XL_KiSS.safetensors
epiCRealism_XL_Pure_fix.safetensors
epiCRealism_XL_V8.safetensors
epiCRealism_XL_V8_KiSS.safetensors
epicrealism_xl_pure_fix.safetensors
epicrealismXL_pureFix.safetensors
epicrealismXL_vxviiCrystalclear.safetensors
epicrealismXL_pureFix.safetensors





