Exécuter en local
unstableinkdream_v80BalanceV3.safetensors
1.99 GB · SafeTensor · fp16 · pruned
Les téléchargements sont servis par Civitai. Certains créateurs demandent d'abord une connexion à Civitai.
À propos de ce modèle
STABLE DIFFUSION 1.5 · Modèles de base
Notes du créateur
anime-Hyper-VAE
我只是把UnstableinkdreamV9.0-anime 加上 HyperSD和VAE,其他沒變,給自己用的,其實可以不需要下載
然後因為hyper-sd
所以建議CFG 1.5~2
取樣方法 LCM
Schedule type Auto
Steps:８-20再多也沒啥意義
anime-Hyper-VAE
I just add HyperSD and VAE to UnstableinkdreamV9.0-anime , no other changes, for my own use, in fact, you do not need to download!
And because of hyper-sd
So we suggest CFG 1.5~2.
Sampling method LCM
Schedule type Auto
Steps:8-20 more steps is meaningless.
如果你喜歡我的工作可以幫我買杯咖啡，謝謝
If you like what I do, you can buy me a cup of coffee. Thanks.
私の仕事を気に入っていただけたら、コーヒーを一杯おごってください。 ありがとうございます。
9.0
It's just a minor tweak to make it more adaptable to all kinds of LORA.
Photo can be transformed into anime by LORA.
anime is also possible.
A few of the LORA I made can be put to good use.
Specialised Negative
只是小幅的微調,讓他更能夠良好的適應各種LORA
photo其實可以靠LORA轉變畫風成為動漫化
anime一樣也可以
幾個也是我做的LORA可以善用
專用的Negative
8.0
it, but I've updated it with LECO and I'm going to put up the photo version first.
Because it is impossible to combine animation and reality, after 7.5 it will be split into three genres - balance/anime/photoreal
But recently things have changed and with some new LECO training methods some of the concepts can be shaved off.
so you can use some lora to chage the style of base model
such as my lora
8.0-balance
2D,2.5D,3D All can do well
But it is better to be in 2.5D digital art or something like that in the direction of fineness, which can give you more surprises.
Better detailing, and can be combined with various LORA variations.
Just Try it
8.5-photoreal version
This is a very realistic model for photorealism and hyperrealism
V7.5-Anime
This is a very well in manga,anime....etc
7.5
the April version of unstableinkdream
Because it is impossible to combine animation and reality, after 7.5 it will be split into three genres - balance/anime/photoreal
7.5-balance
photo anime , movie,manga, photorealism 2D,2.5D,3D
All can do well
Just Try it
7.5-photoreal version
This is a very realistic model for photorealism and hyperrealism ,and is also a bit naughty
V7.5-Anime
This is a very well in manga,anime....etc
V7.3
Using the MBW plug-in to remerge the model more detaile and style photoreal
V7.0
Not uploaded for a long time
Using the auto-MBW plug-in to remerge the model
Compared to V6 this is a more realistic model, but if the prompts are used properly it can also be used to produce anime.
To be honest, the pure models have not improved much since mid-January, most of the models are already adequate, controlling the network and LORA is more important.
The example is a quick draw of 45 cards using wildcrd and dynamicprompts, selected at random.
Why is it a lot of fantasy WAIFU style? I like it ...... But the other styles actually look ok too
I'll leave it to you to try it out
V6
Reconstruction the model to make the prompts itself can be used to create a variety of styles including realistic photos, realistic drawings, and animations like last three fig
model tag like : nvinkpunk,kuvshinov,dreamlikeart,samdoesart and modelshoot style analog style
will effective
and normal tag like : unreal engin 5,octane render, limited palette flat color,Hyperrealism,Photorealism photorealistic also worked well
some negative prompt was come from civitai and huggingface google can find it
V5photoreal- updated
To be honest, I'm sorry, this should come out together with yesterday's V5, one is more towards animation and the other more towards reality
But has not been adjusted very satisfactory, the effect is actually a bit close to V4
If you want to be more photo-realistic, please add (photorealistic) at the begin , and real person's name will effect
If you want to favor animation, please add (nvinkpunk,kuvshinov,dreamlikeart,samdoesart,modelshoot style,) at the end.........not only anime style but girls jump out lol
V5 updated
Enhances cartoonish and stylized intensity with better mastery of the face
model tag like : nvinkpunk,kuvshinov,dreamlikeart,samdoesart and modelshoot style will effective
and normal tag like : unreal engin 5,octane render, limited palette flat color,Hyperrealism,Photorealism also worked well
need nsfw in nagtive prompts
V4 updated
Dont ask me where are v2 or v3,mix a lot of model ,v4 is the best one
mix by V1,RPG,Protogen,dreamlikesamkuvshinov,inkpunk.......
model tag like : nvinkpunk,kuvshinov,dreamlikeart,samdoesart and modelshoot style will effective
and normal tag like : unreal engin 5,octane render, limited palette flat color,Hyperrealism,Photorealism also worked well
UnstableSamInkDream Diffusion
Combined from the following models
A:Unstable PhotoReal v0.5
B:inkpunk
C:dreamlikesamkuvshinov
A0.5+B0.5=D D0.6+C0.4＝UnstableSamInkDream
Trigger Words:
nvinkpunk,dreamlikeart,samdoesart,kuvshinov,
SHA256
045734917B02021598EE924522C55EE27277CC5BA1599B97F7324FEFEF1F3240
ID de modèle ModelTrends.ai
#2036
Signaler le contenu
Rendus d'exemple
Prompts et réglages partagés par les personnes qui ont réalisé ces rendus sur Civitai. Choisissez un rendu pour voir comment il a été fait.
Réglages
Sampler
Euler a
Étapes
35
Guidance
7
Seed
259341533
Taille
512x768
Clip skip
2
Prompt
loong,cloudy sky,claws,eastern dragon,long tail,whiskers,chinese new year,cloud,no humans,gunzan,sky,cloud horizon,sun,outdoors,QINGLONG,SOLO,(Science fiction),((Soaring in the clouds and riding in the fog)),(((The dragon body disappears and appears in the clouds))),(full body:1.3),(ink wash painting:1.2),east asian architecture,<lora:0221 loong_v2:1>,ruanyi0221,snake body,(((floating in the clouds))),, best quality,masterpiece,highres,official art,extremely detailed cg unity 8k wallpaper
Prompt négatif
easynegative, ng_deepnegative_v1_75t, (worst quality:2), (low quality:2),(normal quality:2), lowres, bad anatomy, badhandv4, ((extra limbs)), ((extra legs)), ((fused legs)), ((extra arms)), ((fused arms)), normal quality, ((monochrome)), ((grayscale)), ((watermark)), uneven eyes, lazy eye,bad-hands-5, (((mutated hand))),
Réglages
Sampler
Restart
Étapes
24
Guidance
5
Seed
1252123737
Taille
512x768
Prompt
(leafart,leaf, starry night:1.2),(paper-cut art by Muqi),
(intimidating chimera between a owl, bat, and wolf:1.2),
(Night Sky: black, indigo, violet, silver, and star color theme),
(digital art painting, oil painting, intricate detail, ornate) (trending on pixiv, antialiased, intricate detail, ornate),
<lora:LeafArt_v1_R32:1>
Prompt négatif
(worst quality, low quality:1.5),(text,writing,signature,watermark:1.2),lowres,grainy,blurry,bad drawing,jpeg artifacts,cropped,text,error,extra digit,fewer digits,blurred
Réglages
Sampler
DPM++ 2M Karras
Étapes
50
Guidance
7
Seed
1875486450
Taille
2048x3584
Clip skip
2
Prompt
photorealistic, (4k), (Masterpiece), (realistic skin texture), extremely detailed, intricate, hyper detailed, high resolution, sharp detail, best quality, (reindeer skull), antlers, winter, snow, snowflakes, porcelain, neon lights, red neon, red glowing eyes,
Prompt négatif
worst quality, normal quality, low quality, low res, blurry, text, watermark, logo, banner, extra digits, cropped, jpeg artifacts, signature, username, error, sketch ,duplicate, ugly, monochrome, horror, geometry, mutation, disgusting, bad anatomy, three hands, (extra fingers, deformed hands, polydactyl:1.5), three legs, bad arms, missing legs, missing arms, poorly drawn face, bad face, fused face, cloned face, worst face, three crus, extra crus, fused crus, worst feet, three feet, fused feet, fused thigh, three thigh, fused thigh, extra thigh, worst thigh, missing finger, ugly fingers, long fingers, horn, realistic photo, extra eyes, huge eyes, amputation, disconnected limbs, symmetry, badhandv4, By bad artist -neg, verybadimagenegative_v1.3,
Réglages
Sampler
DPM++ 2M Karras
Étapes
15
Guidance
20
Seed
2546421016
Taille
512x640
Clip skip
2
Prompt
best quality,masterpiece,highly detailed,ultra-detailed,
sports car racing on
cyberpunk 2077,
Prompt négatif
lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name,by bad-artist-anime by bad-artist,(monochrome, greyscale, BW:1.4), (7 fingers:1.9)( 6 fingers:1.9) (twisted fingers:1.9) (too long fingers:1.3) (3 arms:1.9) (3 legs:1.9) negative_hand-neg,
Réglages
Sampler
DPM++ 2M Karras
Étapes
20
Guidance
7
Seed
2780200034
Taille
2048x3584
Clip skip
2
Prompt
photorealistic, (4k), depth of field, (Masterpiece), (realistic skin texture), extremely detailed, intricate, hyper detailed, professional photography, bokeh, depth of field, high resolution, sharp detail, best quality, 1boy, <lora:snow-style-richy-v1:0.7> snowstyle , futuristic city, robotic, snowing, winter, winter clothes, science fiction, mystical portal, deer skull, antlers, man with a deer head, red armor trim, glowing red eyes, moonlight, (night time), stars, contrapposto , aurora, cowboy shot,
Prompt négatif
worst quality, normal quality, low quality, low res, blurry, text, watermark, logo, banner, extra digits, cropped, jpeg artifacts, signature, username, error, sketch ,duplicate, ugly, monochrome, geometry, mutation, disgusting, bad anatomy, three hands, (extra fingers, deformed hands, polydactyl:1.5), three legs, bad arms, missing legs, missing arms, poorly drawn face, bad face, fused face, cloned face, worst face, three crus, extra crus, fused crus, worst feet, three feet, fused feet, fused thigh, three thigh, fused thigh, extra thigh, worst thigh, missing finger, ugly fingers, long fingers, extra eyes, huge eyes, amputation, disconnected limbs, symmetry, badhandv4, By bad artist -neg, verybadimagenegative_v1.3,
Réglages
Sampler
Euler a
Étapes
35
Guidance
7
Seed
2861379811
Taille
512x768
Clip skip
2
Prompt
loong,cloudy sky,claws,eastern dragon,long tail,whiskers,chinese new year,cloud,no humans,gunzan,sky,cloud horizon,sun,outdoors,QINGLONG,SOLO,(Science fiction),((Soaring in the clouds and riding in the fog)),(((The dragon body disappears and appears in the clouds))),(full body:1.3),(ink wash painting:1.2),east asian architecture,<lora:0221 loong_v2:1>,ruanyi0221,snake body,, best quality,masterpiece,highres,official art,extremely detailed cg unity 8k wallpaper
Prompt négatif
easynegative, ng_deepnegative_v1_75t, (worst quality:2), (low quality:2),(normal quality:2), lowres, bad anatomy, badhandv4, ((extra limbs)), ((extra legs)), ((fused legs)), ((extra arms)), ((fused arms)), normal quality, ((monochrome)), ((grayscale)), ((watermark)), uneven eyes, lazy eye,bad-hands-5, (((mutated hand))),
Réglages
Sampler
Restart
Étapes
24
Guidance
5
Seed
882853367
Taille
512x768
Clip skip
2
Prompt
(visionary oil painting by Muqi),
(intimidating chimera between a fox, spider and a humanoid)
(green, brown and and deep yellow colors)
(digital art painting, oil painting, intricate detail, ornate) (trending on pixiv, antialiased, intricate detail, ornate),
<lora:LeafArt_v1_R32:1>,(leafart,leaf, starry night:1.2)
Prompt négatif
(worst quality, low quality:1.5),writing,signature,watermark,worst quality,low quality,normal quality,bad quality,lowres,bad anatomy,disfigured,grainy,blurry,bad drawing,jpeg artifacts,cropped,bad hands,bad fingers,text,error,missing fingers,extra digit,fewer digits,malformed,mutated,blurred
Réglages
Sampler
DPM++ 2M
Étapes
20
Guidance
7
Seed
3272910219
Taille
512x512
Clip skip
2
Prompt
imcc style, orient princess, portrait, focus on face, fantasy, color pencil drawing, <lora:Iain_McCaig_style:1.0>
Prompt négatif
watermark, signature, frame, border,
Modèles similaires
Modèles d'IA pour générer image, vidéo et texte
unstableink8 est un modèle de base pour la famille Stable Diffusion 1.5, référencé sur ModelTrends.ai, un catalogue en lecture seule de modèles d'IA open source pour l'image, la vidéo et le texte. Comparez-le aux autres modèles Stable Diffusion 1.5, consultez son score de chaleur pour suivre sa tendance et ouvrez-le sur PirateDiffusion ou BitVector pour l'essayer.
Plus de modèles Stable Diffusion 1.5Plus de Modèles de base Stable Diffusion 1.5drawingspaintingsbaseartstyleskuvshinovdreamlikeartnvinkpunkbase model
Parcourir par famille de modèles
Modèles Text / LLM
411
Modèles Anima
509
Modèles Chroma
14
Modèles Flux
1,071
Modèles Flux 2 / Klein
1,237
Modèles MiniMax H3
331
Modèles Hunyuan
197
Modèles Ideogram
20
Modèles Krea2
674
Modèles Qwen
17
Modèles Qwen2
36
Modèles Ltx2
150
Modèles Stable Diffusion 1.5
6,465
Modèles Stable Diffusion XL 1.0
15,432
Modèles Zimage
804
Modèles Wan
919
Model
Trends
.ai
© ModelTrends.ai
|
Fabriqué au Japon
|
© 2026