SUMMARY: Excellent artist style reproduction with vast knowledge base, excellent prompt understanding/adherence and also improved support of natural language.
It is highly recommended to use this model with artist tags, style tags, or at least stylistic LoRas as the base style is very neutral.
LoRa suggestion: https://tensor.art/models/896574003475409210
----------------------------------------------------------------
10/11/2025: If you're looking for a way to increase CFG for more stability and cleanliness without blowing out the colors I highly suggest trying out this little friend here: https://tensor.art/models/879893168689033646 Put it on -0.2 and you can increase CFG by 1.5-2 or even higher, also Euler A becomes much more stable. Or you can put it on +0.1-0.2 and have more vibrant contrast on lower CFG. Very useful!!
----------------------------------------------------------------
11/08/2025: Thank you all for contributing your great art with this model and for all the new people here I strongly encourage you to scroll further down the gallery as it depicts just amazing sutff. Truly a source of inspiration! (V2 & V1)
----------------------------------------------------------------
V2: Improved hand stability, slight adjustments to facial structure, eyes, hair.
----------------------------------------------------------------
-----
Hey, this is my first mix! Give it a try! Updates will come soon too. This merge is build on the GENESIS LoKr model, Noobai and NTRmix XIII. GENESIS is a big extention to NoobAi's and Illustrious artist and style base, which also supports Niji styles.
Here is a list of new concepts that are supported by GENESIS:
Artist Expansion:
https://huggingface.co/rim0/GENESIS/blob/main/Art%20StyleV3.txt
Nji Style Tags:
https://huggingface.co/rim0/GENESIS/blob/main/niji%20StyleV3.txt
Lighting Effect Tags:
https://huggingface.co/rim0/GENESIS/blob/main/Lighting%20Effect.txt
And of course everything what worked for noob and illustrious will work here too.
SETTINGS:
Steps: 20-30 (I mostly use 25, but 28-30 gives better eyes without adetailer)
cfg : 2-6 (I mostly use 4)
Sampler: dpm++ 2M SDE GPU and SGM Uniform (Euler a or dpm++ 2M and SGM Uniform if you use the other negative prompts listed below)
HiresFix: ESRGAN_4x / 10 steps / 0.3 denoise
RESOLUTIONS:
832 x 1536
768x1344
832 x 1216
896x1152
1024 x 1024
1152x896
1216 x 832
1344x768
1536 x 832
Positive Prompt
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest
Negative Prompt
Base option for more flat hand drawn styles:
worst quality,bad quality,jpeg artifacts,(bad hands,bad finger,anatomy),(artist name,copyright name, watermark, text,english text),(bad feet,bad toes)
Enhanced performance for 2.5d (dpm++ 2m SGM Uniform
worst quality, low quality, bad anatomy, bad hands, bad feet, extra digit, fewer digits, watermark as default, add low quality artists. Example) worst quality, low quality, bad anatomy, bad hands, bad feet, extra digit, fewer digits, watermark, chihuri, bb (baalbuddy), iesupa
For a softer look in 2.5d (dpm++ 2m SGM Uniform)
lowres, (bad), text, error, fewer, extra, missing, worst quality, jpeg artifacts, low quality, watermark, unfinished, displeasing, oldest, early, chromatic aberration, signature, extra digits, artistic error, username, scan, [abstract], multiple views, censored,
I also like to combine the flat style of some artists with the 2.d adjusted negative prompts and dpm++2m SGM Uniform to give the flat style more plasticity.