A quiet gaze, suspended between reality and imagination.
Reijality v3
Reijality v3 is a gentle evolution of v2.
The model keeps the same visual identity, but the gaze feels clearer and more present. Skin rendering and lighting are slightly more refined, giving portraits a subtle cinematic touch.
This update is not meant to change the model’s character, but to refine it.
Changes in v3
clearer gaze and eye focus
slightly refined skin rendering
improved light response in portraits
same visual identity as v2, with slightly clearer focus and presence
☺ sweetspot example:
| + | masterpiece, best quality, film still, cinematic lighting, studio lighting, rim lighting, soft shadows, volumetric light, subsurface scattering
| —| worst quality, low quality, lowres, blurry, jpeg artifacts
Euler a
30 -34 steps
CFG 4.8 - 6.8hires x 1,2 / steps 10 / denoising 0,1 / CFG 1
Reijality v2
Naturally unnatural ~ that is Reijality. Not the decision between something, but the conscious choice for both at the same time. It's not an either/or game of dominance and surrender. No part gets lost in the other ~ instead, both stand next to each other on an equal footing, interwoven, merged, but not swallowed up. Light plays on skin ~ not just visible, but noticeable. A shadow play that not only contours, but also entices. Visible and yet hidden. As if it wanted to be discovered ~ but only by those who really WANT to look. And even then: just a touch.
These sensual moments do not thrive on openness, but on the in-between. From hinting. From the game. About breathing between light and dark. About whispering between the colors. Of touch that only becomes noticeable in the imagination. Unreally real. Unnaturally natural.
Reijality v1
hey guys, ☺
do u also like playing with the inbetween like I do?
Step into a world where dreams take shape, where whispers of magic weave through every brushstroke, and where fantasy is no longer just imagined ~ it’s illuminated. ✨
Welcome to Reijality, a realm born from wonder and artistry, designed to breathe life into the ethereal, the magical, and the beautifully surreal. Here, creation is a portal to the extraordinary ~where elegance meets mystery, where light holds secrets, and where each gaze invites u deeper into an enchanted universe. Can u hear it whispers?
This is a place where fantasy blossoms into reality, where the unseen becomes tangible, and where imagination knows no bounds. It captures possibility. Are u ready to wander, to dream, and to embrace the magic that lingers in every shadow and shimmer?
🌙✨ Welcome to Reijality. Welcome home. ✨🌙
Let ur creativity guide u and uncover the stories waiting to be told. ✨
▼ recommended settings examples V1:
Euler a / DDPM30 stepsCFG 5-8Clip skip 2+prompt example
best_quality, absurdres, ultra-detailed, highly_aesthetic, 8K, UHD, highly_detailed_eyes, depth_of_field, subject_focus, or
masterpiece, high_quality, highres, sharp_focus, detailed_eyes, - prompt example
worst_quality, bad_quality, poorly_detailed, young,extra_fingers, malformed_hands, deformed_feet,I am happy if u enjoy and stay curious abt ur story tellings ✨
If u like what I do consider leaving me a like or a comment, much appreciated and helps me getting a feeling for tinkering ~♥
Description
VAE baked in
v2 tinkered with
FAQ
Comments (15)
V2 is fire! ❤️❤️❤️
very nice realistic model for IL. I got some issues with proper licking or deep throat or cum details, but so far my first IL model to have some nice results.
Hi there!
I´ve had the same Issues, and solved it with the following Promt-Snippets:
- for cum:
((detailed cum, best quality cum, very thick cum, very glossy cum, excessive cum)), ((cum on face), (cum in open mouth), (cum on body), (cum in pussy), (cum drippings), (cum strings))
- for "salvia, drool, pussyjuice, pussy-juice, pussy juice", it doesn´t work properly...but "clear liquid" does O.o
((detailed clear liquid, best quality clear liquid, very thick clear liquid, very glossy clear liquid, excessive clear liquid)), ((clear liquid on mouth), (clear liquid on penis), (clear liquid on pussy), (clear liquid drippings), (clear liquid strings))
- for the "deepthroat":
((irrumatio, deepthroat):1.5)
I hope, this helps ;-)
Cheers,
D.
thank you for this model, what is the best recommended picture size?
Thanks for ur question!
In general, I recommend working around 832 × 1216 (or similar formats like 720 × 1280).
U can also try 1040 × 1510, though that one is more experimental.
Just keep in mind: with more stretched-out image sizes, u might occasionally get an extra finger or two popping up here and there. 😄
best regards ~☺ Reij
How do you describe two actors? one being male and the other being female?
Following this thread. Been wondering the same.
hey you two ☺
This is fundamentally a challenge at times.
It can be helpful to divide the input prompt or use ( ) to set the appropriate descriptions. It doesn't always work down to the last detail. Another option is to increase the CFG, but here too, depending on the prompt, a bit experimental, sometimes ( ) are enough, sometimes accordingly with strengths in the brackets to emphasize individual passages (blonde hair:1.2) e.g. or higher or lower depending on what you want to emphasize ☺
reijlita Got it! Thank you! Btw. I've made close to 400 renders with Reijality v2. Without a doubt THE best checkpoint on this entire site. Thought you'd wanna know lol.
VerseVisualsAi thank u ☺ kind that u said that :) I'm very pleased that it worked. and if you like the model ^^
I know I'm late to the game here, but in addition to reijita's suggestion, sometimes providing a name to each character can help when you put their descriptions in parenthesis "()" or brackets "[ ]". (PLEASE NOTE: When using parenthesis, remember they amplify whatever is between them. Brackets usually lessen the impacts of whatever is between them. You can use this to balance out the characters as needed.
If it is a proper character name, so much the better, since a lot of checkpoints have the appearance of major characters buried in the base model.
Another option is using LoRA. This will help strengthen characters, but I would suggest toning down the weights as one LoRA may dominate the other. As before, seperate the trigger words in their respective descriptions.
For Example: (See Result at Example of Two Characters (Male/Female) | Civitai
1boy, 1girl, side view, dynamic angle, couple intimate, affection, BREAK
[1girl, candi looking at 1boy
long hair, blush, blue_eyes, sexy_eyes, blonde_hair, large_breasts, wavy_hair, freckles, curly_hair, Pink turtleneck sweater dress, long sleeves, short dress, white boots, thighhigh boots], BREAK
[1boy, looking at 1girl, Harry_Potter, black_hair, dark_brown_eyes, grey_hoodie, dark_denim_jacket], BREAK
semirealistic, realistic, low contrast, (anime:0.4), diffused light, sidelighting, frontlighting, bright, soft lighting,
In this example, I used Harry_Potter. That isn't necessarily what I am looking for, but I do know that is enough of an established character to influence the separation. In the attached post, you will note that this was done with a prompt and one embedding. No LoRAs were used.
I hope this helps :)
@pope_phred Thank you very much for your detailed suggestions and the different options you listed :) that's very kind. I would also like to thank you for making me smile, the conclusion of your collection of ideas was unexpected ^^
It is both reality and better than reality at the same time. I'm in love.
thank u very much 4 ur kind words :) happy if u enjoy ☺
I got some skin details, that is very close to some 3wolf models.
Details
Files
Available On (3 platforms)
Same model published on other platforms. May have additional downloads or version variants.