Update April 28th Added version V4 pruned model.
A model with professional RPG industry artists in mind. Those working in video games, board and tabletop games as well as concept art and book covers should get good use from this model. The new version, V4, packs in more training for creatures and RPG subjects including a more illustrative style.
Detailed tutorial on how I get the results in the preview images.
Check here if you're having trouble getting the same results. Initial generation size and VAE usage is key.
You can prompt any style you need with these models, but the default aesthetic is listed for each of the models in this handy list.
Different models available, check the blue tabs above the images up top:
Stable Diffusion 1.5 (512) versions:
V4 inpaint Inpainting version of V4 that's good for outpainting.
V4+VAE Same as V4 but with the added convenience of having a preset VAE baked in so you don't need to select that each time. Saves on vram usage and possible NaN errors. Speeds up workflow if that's the VAE you're going to use anyway.
V4 New aesthetic added to make it a bit more illustrative. More training for better coherency for all things RPG.
V3+VAE Same as V3 but with the added convenience of having a preset VAE baked in so you don't need to select that each time. Saves on vram usage and possible NaN errors. Speeds up workflow if that's the VAE you're going to use anyway.
V3 Stronger painterly style. High contrast and sharpness. Even more RPG knowledge.
V3 inpaint Inpainting version of V3 that's good for outpainting.
V2 Stronger painterly style. High contrast and sharpness. More RPG knowledge.
V2 offset Noise Offset added making more contrast and bringing the model back to photoreal.
V2 Art Trained model. Very artsy. Strongest painterly style. Less details and bigger brush strokes to mimic digital painting style pre-AI.
V2 inpaint Inpainting version of V2 that's good for outpainting.
V1 Smoother renders with least painterly effect.
V1 inpaint Inpainting version of V1 that's good for outpainting.
Stable Diffusion 2.1 (768) versions:
SD 2.1 768 V1 Strong painterly style, very coherent with hands and objects. Higher native resolution and detail. Not good for nudity.
Not as effective, but here's the LoRA if you like to use that instead:
A to Zovya RPG Artist's Tools LoRA
Do you have requests? I've been putting in many more hours lately with this. That's my problem, not yours. But if you'd like to tip me, buy me a beer. Beer encourages me to ignore work and make AI models instead. Tip and make a request. I'll give it a shot if I can. Here at Ko-Fi
Description
FAQ
Comments (19)
The best models ever, doesn't mean others are bad, no no, but THIS is special, my special. thx so much for this Zovya
Doesn't work for NMKD Stable Diffusion. anyway, you can look into this.
i was trying to generate this image with Zovya_RPG_Artist_Tools_V2_Offset model.
https://civitai.com/images/436864
i don't know what i'm doing wrong. i can't generate the same result using same settings.
what i generated
https://i.imgur.com/bNF6Ixg.png
my settings - https://i.imgur.com/edcn9am.png
there's a link in the description that shows a technique I use to get a sharper, high contrast image like those in the preview. try that if that's what you mean. otherwise, it still won't be EXACT since all GPUs are different. but if you mean the effect, that's how I get it.
@Zovya thank u for replying.
i generated these using tesla t4 on colab pro.
are you using some embeddings when generating. cuz even the character pose is different from the original one.
i have generated preview images with other models but they weren't much different from the original images
@Snow_Maan no embeddings
There is a setting in AUTO1111, I believe it is the `Random Number Generator source` & some samplers in ComfyUI. CPU in general will produce similar or the same result each time, where GPU or NV will vary & be dependent on often that specific GPU, as Zovya mentions. Not sure how the colab's work, or if you get access to those options.
This model comes up with a filter of "trained".
Usually, trained models have details like "epoch" count, stuff like that. Woud you mind adding that info please?
there's no way to know at this point, this model has been finetuned for months
I just went through my hundreds of 1.5 fantasy/cg models to rank the best and cut the worst, and this one is 100% top 5, maybe best of all. I only looked at v3 so far, because I hesitated at the unpruned size, but if ever a model was worth that size commitment in 1.5, it's this one. Came back for v4, pruning be damned (although if you did I would worship at your feet).
Can you recommend some other ones?
This one is great, but boy does it have trouble generating armor other than plate, and still does a lot of bare chests/legs.
@norber4444558 Dreamshaper is probably the best regarded, any of the Duchaiten models are some of my favorites, Copax models are great, RealDream models are the best in flexibility, the Xeno models are very good, RealCartoon, the list really goes on and on. I have hundreds of fantasy/render models that I like, and the majority of models on this site fall into that category. Hope you find what works for you!
Difference between v4 and v4 vae models? What vae is used here? Will it make a difference in output?
Any chance of this being available for on-site generation again? It worked fantastically with one of my models, and I would love to continue working with it.
Bro why i can't generate this model onsite anymore? Pls I love this model so much!
Will there be pruned version of v4?
v4 is a downgrade sadly. All the female faces are very similar and look super young. the style feels more generic, and over all I feel V4 goes in a worse direction.
bro we need pruned version, thanks for your hard work
It's impossible for me to generate a male with hair other than the slicked back hair with bangs, I don't understand why it's not picking up ANY prompts about hairstyle I put in, whether it's v4 or v3.
Any suggestions?
Otherwise, great for generating D&D characters! Would love to see an XL version.