Pony V8
Recommend
Trigger: score_9,score_8_up,score_7_up,score_6_upSteps : 20
CFG : 4
Sampling : euler a
Schedule type: simple
Nprompt:
safe_neg, worst quality, bad quality, low quality, missing fingers, extra fingers, rembrandt lighting, darkness, night, sweat, dutch angle,
Pony V7 (3D Real)
Recommend
Trigger: score_9,score_8_up,score_7_up,score_6_upSteps : 40~50
CFG : 6
Dimension: 680 x 1024, 768 x 1024
You can also use After Detailer for face
Sampling : euler a
Schedule type: simple
Nprompt:
muscular, fury, child, kid, easynegative, distorted fingers, distorted hands, aliasing, distortion, lowres, muscular, realistic, downsampling, poorly drawn, logo, monochrome, watermark, low-quality, bad-anatomy
DreamyIdol V1 (Illustrious)
Steps : ~30
CFG : 4.5~7
Dimension: 680 x 1024, 768 x 1024, 896x1152,768x1344
You can also use After Detailer for face
Sampling : euler a
Schedule type: simple
Nprompt: lowres, blurry, bad anatomy, bad hands, bad eyes
__________________________________________________________________
Pony XL V5 3DReal
Combination of 3D and real. The face is slightly round and the hair has high detail.
Note: Schedule type: SGM Uniform for better result
Recommend
Trigger: score_9,score_8_up,score_7_up,score_6_upSteps : 20~30
CFG : 7~8
Dimension: 680 x 1024, 768 x 1024
You can also use After Detailer for face
Sampling : DPM++ 2M SDE
Schedule type: SGM Uniform
Nprompt:
score_6, score_5, score_4, ugly face, low res, blurry face, black and white, muscular female , extra penis
Pony XL V4
Adjust the facial appearance and body proportions, with more realistic background details.
Details are more about photography than painting.
Recommend
Trigger: score_9,score_8_up,score_7_up,score_6_upSteps :30
CFG : 7~8
Dimension: 680 x 1024, 768 x 1024
You can also use After Detailer for face
Sampling : DPM++ 2M SDE
Schedule type: Karras
Nprompt:
score_6, score_5, score_4, ugly face, low res, blurry face, black and white, muscular female , extra penis
Pony XL V3 (Couple)
Details are more about photography than painting compared to the V3(single) version
Recommend
Trigger: score_9,score_8_up,score_7_up,score_6_upSteps :30
CFG : 7
Dimension: 680 x 1024, 768 x 1024
You can also use After Detailer for face
Sampling : DPM++ 2M SDE
Schedule type: Karras
Nprompt:
score_6, score_5, score_4, ugly face, low res, blurry face, black and white, muscular female , extra penis
Pony XL V2
Hello everyone,
I am thrilled to share with you the latest update, Copax Pony XL V2!!!
This version brings several changes:
Significantly reduces distortion errors when creating human images
Character Appearance: Refreshed and enhanced to be more.
Skin Details: Adjusted skin details, from texture to how light reflects off it.
Background: The scenery has become richer and more detailed.
Improved hand details.
If you enjoy the work I’m doing and want to see more updates in the future, please consider supporting me by leaving a like, review, tip, or making a donation. Your support will be a great motivation for me to continue bringing new models in the coming times.
Thank you so much for always being there and supporting me!
Recommend
Trigger: score_9,score_8_up,score_7_up,score_6_upSteps :30
CFG : 7
Dimension: 680 x 1024, 768 x 1024
You can also use After Detailer for face
Sampling : Euler a
Nprompt:
score_5, score_4, bad hands, bad eyes, bright pupils, white pupils
_____________________________________________________________________
Copax_Pony_XL
The model has been improved to receive better prompts and reduce error rates when creating images
However, sometimes you may still encounter some errors when creating images with complex details
Recommend
Trigger: score_9, score_8_up, score_7_up,Steps :30
CFG : 7
Dimension: 680 x 1024
You can also use After Detailer for face
Sampling : DPM++ SDE
Nprompt:
score_6, score_5, score_4, (worst quality:1.5), (low quality:1.5), (normal quality:1.5), lowres, bad anatomy, bad hands, multiple eyebrow, (cropped), extra limb, missing limbs, deformed hands, long neck, long body, (bad hands), signature, username, artist name, conjoined fingers, deformed fingers, ugly eyes, imperfect eyes, skewed eyes, unnatural face, unnatural body, error, painting by bad-artist
Description
FAQ
Comments (8)
I just ran a series of various tests to measure 13 models against each other. The test used a scoring system of 1-5 (1 being the lowest). I tested prompt adherence, accuracy of the picture itself, ability to work as a refiner, detail and complexity of scenes, quality of final render, and several other little things.
I tested your V3 Single model.
This model scores third place, which after using it for a while, I'm not surprised due to the quality of it. It even beat out your latest model in almost every category I scored it on. Where the model shines was fairly good scenery complexity, the extremely high marks for anime quality, accuracy of the image, and extremely high marks for creating quality final images.
The model is universally works great as a refiner, and as the main model for further refinement with other models, having the most consistent quality across the board. This is probably due to scenery complexity, and accuracy of the prompt and images.
Don't sleep on this one folks. It is truly a fantastic model.
Unfortunately, your latest V4 model scored a tad lower compared to the rest (but not the lowest by a long shot). It suffered from loss of scenery complexity, and dynamically and random generated faces - they were nearly all the same.
That being said, of the 50 or so models I tested prior to this more detailed testing, these two made it to the top of my list of 13, and both scored excellently.
I'm very happy to read your comment. Thank you for your support; I will take your feedback into account to continue improving the model’s quality in future updates.
@Copax cheers mate. I made some alterations to the original post. I was a bit tired when I wrote. Seriously, great models. Keep up the excellent work!
Have you shared your model rankings somewhere, and if not, would you be willing to? :)
@randomuser_mon I'm not sure I'd be comfortable doing that. Reasons being:
- I don't want to start drama or dissuade folks I feel are doing a wonderful job putting models together, mixing them, etc.
- I don't want to upkeep it, since it would be outdated rather quickly.
- I definitely don't want people to judge things based upon my specific opinions.
- It is preferable if people don't question my methodology, as I'd have to update it, there would likely be debate about it, etc., and some of that is very subjectively based.
I only let a few folks know what I feel were the highest ranking of the bunch, from my perspective. I'm not even sure I'll do it again.
@TurinBjorn Alright, no problem!
thanks, i looked at the examples, and guessed V4+V3 are the best but couldn't decide which one. Now i picked V3 sinjce you wrote its better than V4. I cant keep multiple version of same Model since there already dozens of model out to be collected ;).
@pink0909 The "single" models seem to be the best in my opinion, but that's very subjective. V3 was the one I used in my initial assessment, but they've gone further down the road with these, and I haven't updated my review. I often utilized it as a base, and tweaked from there, since it's prompt coherence was so good, along with the posing, complexity of scenes, etc.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.