Run Locally
RussianDollV3.safetensors
144 MB · SafeTensor · fp16 · full
Downloads are served by Civitai. Some creators ask for a Civitai login first.
About This Model
STABLE DIFFUSION 1.5 · LoRAs
The RussianDoll LoRA is used for creating the likeness of non-specific beautiful Caucasian women. This V3 update works better in Realistic and Semi-Real 2.5D, however sometimes generates cleft chins. We have accounted for this in the included tokens, but may still occur when guidance is set above 7. It is also advised to use a lora weight lower than 0.4 combined with a base of for best results. The LoRA has good reviews and works well, but there are notable comments about the Russian authenticity of the look, so you may also want to add specific characteristics or other geolocations and put [asian, slavic, etc] in negative prompts when using Chilloutmix as your base. For an alternative look, also try
Creator notes
This is my first Lora, please be nice and forgiving for any mishaps.
RussianDollV3
After being inspired by the Korean Doll Likeness by Kbr, I wanted to make one for more beautiful caucasian faces. Since I am deeply impressed by the Russian's high cheekbones, I decided to make a Lora to imitate this type of face. I know high cheekbones are not exclusive to Russia, but their models have a high percentage of this trait so I named it as such.
I trained this Lora with 28 images at the beginning, using a combination of (beautiful model, fit faces, and high cheekbones) in my prompts. No one or person was used as training data. From those 28 images, I made V1, which was unsatisfactory so I trained again with 100 steps, 8 epochs, and 6 batches to get V2. Using V2, I generated 24 more pics which I used to train for V3, the current model.
By no means is this model perfect, but it does get the high cheekbone that I find really attractive and sexy. There are two minor flaws: it is biased toward NSFW, and sometimes it generates a cleft chin. Just input them into the UC or add "SFW" and "sharp chin" in the prompt
Tips:
• The Lora works even better when there are (beautiful model and fit faces) in the prompt.
• I find the best result around 0.25~0.4. Don't go above 0.5 as it starts to get weird. The lora default is , remember to change it accordingly.
• This model works best in ChilloutMix. It also works decently well in 2.5D.
Disclaimer:
As the author, I regret to inform you that the LORA model is not intended for commercial use. This decision has been made in order to prevent any potential controversies or legal issues related to impersonation or defamation that may arise from the use of the model for commercial purposes. Furthermore, allowing commercial use of the model may have a negative impact on the development of the AI community. Therefore, I strongly advise against any commercial usage of the LORA model.
Trigger words
1girl
russian
SHA256
BF1AEF0B6348FD5C96FC0701C4D1BDD743D7A53B3CCC7CE83AD494441E7208B7
ModelTrends.ai Model ID
#463
Flag content
Example renders
Prompts and settings shared by the people who made these renders on Civitai. Pick a render to see how it was made.
Settings
Sampler
DPM++ SDE Karras
Steps
27
Guidance
7
Seed
3324874028
Size
576x912
Prompt
Russian, <lora:russianDollLikeness_v3:0.5>, (best quality, ultra high res, raw photo, masterpiece, extremely detailed photograph), (anime, photorealistic:1.4), 1girl, solo focus, sharp chin, looking at viewer, dynamic angle, black glistening leather jacket, <lora:eyeLora_eyesV10:0.4>, gray hair, dark blue eyes, from above, (ponytail, wavy hair)
Negative prompt
(painting by bad-artist, bad_prompt:0.5), (low quality, worst quality, lowres:1.5), (nipples:2), skin spots, extra limbs, bad hands, jpeg artifact, (see-through:2), bad eyes,
Settings
Sampler
Euler a
Steps
30
Guidance
7
Seed
3203383375
Size
512x512
Prompt
Russian, chubby cheeks, stewardess, serious face
Negative prompt
easynegative, Asian-Less-Neg, bad-picture-chill-75v, kkw-Extreme-Neg, ng_deepnegative_v1_75t, verybadimagenegative_v1.3
Settings
Sampler
DPM++ SDE Karras
Steps
27
Guidance
7
Seed
1527191512
Size
512x768
Prompt
(Russian), <lora:RussianDoll6:0.35>, (photorealistic:1.4), 1girl, (sharp chin:2), (snow background, snowing, forest) (Flap Pocket Fuzzy Trim Hooded Puffer Coat, fully clothed, very thick winter jacket, thick jacket with fur hoody:1.3), solo focus, upper body, looking at viewer, close up, <lora:eyeLora_eyesV10:0.4>, (blonde hair, twintail hairstyle), blue eyes, (wavy hair), sfw, detailed background, bokeh, (best quality, ultra high res, raw photo, masterpiece, extremely detailed photograph),
Negative prompt
(painting by bad-artist, bad_prompt:0.5), (low quality, worst quality, lowres:1.5), (nipples:2), skin spots, extra limbs, bad hands, jpeg artifact, (see-through:2), bad eyes, bodysuit
Settings
Sampler
DPM++ SDE Karras
Steps
27
Guidance
7
Seed
4208511369
Size
768x512
Prompt
Russian, <lora:RussianDoll2:0.4>, (best quality, ultra high res, raw photo, masterpiece, extremely detailed photograph), (photorealistic:1.4), 1girl, (nudes), light smile, solo focus, upper body, looking at viewer, close up, pool, half-submerged in water, <lora:eyeLora_eyesV10:0.4>, auburn hair, blue eyes, from above, (wavy hair)
Negative prompt
paintings, sketches, (worst quality:2), (low quality:2), (normal quality:2), lowres, normal quality, ((monochrome)), ((grayscale)), skin spots, acnes, skin blemishes, age spot, glans, (painting by bad-artist, bad_prompt:0.5),
Settings
Sampler
DPM++ SDE Karras
Steps
29
Guidance
7
Seed
1497207883
Size
528x776
Prompt
Russian, model, fit face, <lora:RussianDoll6:0.3>, (best quality, ultra high res, raw photo, masterpiece, extremely detailed photograph), (photorealistic:1.4), 1girl solo focus, looking at viewer, pov, closeup, (sharp chin:1.5), <lora:eyeLora_eyesV10:0.4>, (ponytail), auburn hair, green eyes, from above, young woman, (navy blue business suit:1.5), (office room background, indoor),
Negative prompt
(painting by bad-artist, bad_prompt), (low quality, worst quality, lowres:1.5), skin spots, extra limbs, bad hands, jpeg artifact, (see-through clothing:1.5), cleft chin
Settings
Sampler
DPM++ 2M SDE Karras
Steps
24
Guidance
4
Seed
369396530
Size
512x768
Prompt
candid photograph of (woman and man together) <lora:RussianDoll:0.4> russian wearing a turtleneck sweater and fur jacket and wool skirt ((embracing)) Ukrainian President <lora:Zelenskyy2:0.7> zelenskyy,(3/4 shot),(slight smirk),Kilo class submarine in the background,sunglasses,((Ushanka)),Ukrainian flag,(blue and yellow flag),((Kyiv))
Negative prompt
breasts,topless,nipples,nude,nsfw,cleavage,(watermark),easynegative,black and white
Settings
Sampler
DPM++ SDE Karras
Steps
27
Guidance
8
Seed
3324874026
Size
512x910
Clip skip
2
Prompt
Russian, (best quality, ultra high res, raw photo, masterpiece, extremely detailed photograph), (realistic, photorealistic:1.4), 1girl, solo focus, sharp chin, looking at viewer, dynamic angle, black glistening leather jacket,red hair, dark blue eyes, from above, (ponytail, wavy hair) <lora:eyeLora_eyesV10:0.3> <lora:russianDollLikeness_v3:0.7>,
Negative prompt
(painting by bad-artist, bad_prompt:0.5), (low quality, worst quality, lowres:1.5), (nipples:2), skin spots, extra limbs, bad hands, jpeg artifact, (see-through:2), bad eyes,skin spots
Settings
Sampler
DPM++ SDE Karras
Steps
29
Guidance
7
Seed
1497207883
Size
528x776
Prompt
Russian, model, fit face, <lora:russianDollLikeness_v3:0.3>, (best quality, ultra high res, raw photo, masterpiece, extremely detailed photograph), (photorealistic:1.4), 1girl solo focus, looking at viewer, pov, closeup, (sharp chin:1.5), (ponytail), auburn hair, green eyes, from above, young woman, (navy blue business suit:1.5), (office room background, indoor),
Negative prompt
(painting by bad-artist, bad_prompt), (low quality, worst quality, lowres:1.5), skin spots, extra limbs, bad hands, jpeg artifact, (see-through clothing:1.5), cleft chin
Similar Models
AI Models for Image, Video & Text Generation
russiandoll is a LoRA for the Stable Diffusion 1.5 family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other Stable Diffusion 1.5 models, check its heat score to see how it is trending, and open it on PirateDiffusion or BitVector to try it.
More Stable Diffusion 1.5 modelsMore Stable Diffusion 1.5 LoRAsrealisticsemi-realwomenanatomywhitephotorealisticwomanrussian
Browse by Model Family
Text / LLM Models
411
Anima Models
509
Chroma Models
14
Flux Models
1,071
Flux 2 / Klein Models
1,237
MiniMax H3 Models
331
Hunyuan Models
197
Ideogram Models
20
Krea2 Models
674
Qwen Models
17
Qwen2 Models
36
Ltx2 Models
150
Stable Diffusion 1.5 Models
6,465
Stable Diffusion XL 1.0 Models
15,432
Zimage Models
804
Wan Models
919

Model
Trends
.ai
© ModelTrends.ai
|
Made in Japan
|
© 2026
