Krea 2: simple gen workflow for high quality realism + lots of info & tips — krea2_simple_v1

All the images attached to the post were generated directly with this workflow with no further editing.

Nodes & Models

Custom Nodes:

RES4LYF - A very popular set of samplers & schedulers, and some very helpful nodes. These are needed to get the best outputs, IMO.

RGTHREE - ( Recommended ) A popular set of helper nodes. If you don't want this you can just delete the seed generator and lora power loader nodes, then use the default comfy nodes instead. RES4LYF comes with seed generator & lora nodes as well, I just like RGTHREE's more.

ComfyUI GGUF - ( Optional ) Lets you load GGUF models, which for some reason ComfyUI still can't do natively. Once installed, you use the "Unet Loader (GGUF)" node to load the model. If you're not using any GGUF models you can just skip this.

Required Models:

Important Note: If you can, you should use the Int8 Convrot version of the model (unless you want higher quality using BF16). The Int8 Convrot model is almost 2x as fast to gen with, and is the same quality as FP8. Massive free speed boost. You will need to update your ComfyUI, support was only added early July 2026. You will also need an NVIDIA GPU, and CUDA version 130 or higher.

Main model: Krea2 RAW B16 / FP8 / Int8 Convrot | or | Krea2 RAW GGUFs - It is strongly recommended that you use the RAW main model with the turbo lora at 0.6 strength instead of the Turbo main model when making photo-real images. It gives WAY better results, and the only downside is that it takes a bit longer to gen. Gen times are already pretty short, so that's not a big deal.

The main workflow assumes you're using the RAW model with the turbo lora, and the settings will be very bad if you use the turbo main model instead.

Even the 'full_turbo' workflow still uses the raw model, seeing as you can just set the turbo lora to 1.0 strength and then it does pretty much the same thing as the turbo main model.

Turbo Lora: Rank 64 Turbo Lora - Using this with the RAW model at ~0.6 strength is better than using the Turbo model. The only real downside is speed. Even then, if you're in a hurry you still have the option of upping the strength to 1.0, which makes it just like the turbo model. Gen times are only 50% longer when it's at 0.6, so it's not really worth it to use full turbo IMO.

Anti-Censorship Lora: 2 Vector Bypass Lora - You should use this even if you're doing SFW stuff. More detail is below, but essentially this will massively improve prompt adherence, facial expressiveness, character detail, and numerous other things. There is no downside as long as your sampler settings are good (which this workflow takes care of for you). Do not use other bypass loras , they go too far or cause degradation of quality; this is the only one that works properly.

Text Encoder: Qwen3 VL 4B - Use the BF16 one if you can. Some people say text encoder quality doesn't matter much & to use a lower sized one, but it does matter and it affects quality.

If you're using a GGUF text encoder for some reason, swap out the "Load CLIP" node for a "ClipLoader (GGUF)" node.

VAE: Wan 2.1 FP32 VAE - This gives you sharper, clearer images than when using the Qwen Image VAE. There is no downside. It works because the Wan & Qwen Image VAEs are almost identical, and the FP32 precision improves the quality.

There is an alternative VAE you can use that's even sharper, but it has drawbacks so I've detailed it in the info section further down.

--

That's the end of the required info, you can stop here if you just want to use the workflow!

--

Optional Turbo Finetunes (updated 2026-09-18)

I don't recommend any turbo finetunes for SFW realism yet. Not that I've tested them all or anything, there may be a good one out there I just haven't tried.

But I can recommend Muse by Stable Yogi for NSFW realism. It's the only finetune I've tried that makes images on-par for quality with the raw model. Use euler/simple 8 steps with the full turbo workflow from this post.

Compared to RAW, it's less creative - but that may not matter much to you depending on what NSFW thing you're going for

Compared to RAW, it's really bad at SFW because it tends to insert NSFW stuff into SFW prompts, so avoid using it for SFW stuff

I recommend See-Krea2 for anime-style images. It's SFW & NSFW capable, and doesn't really have any downsides compared to the raw model. It can only do anime and anime-adjacent stuff, but it's damn good at it. There are some good positive prompt tags for it in the non-realism section below.

Info & Tips

Alternative Sharpening VAE

The Wan 2.1 Upscale2x VAE gives you even sharper images than the Wan FP32 VAE (it's VERY noticeable), but it sometimes introduces extra artifacts into the image and it also amplifies existing ones. It's up to you whether you think it's worth it or not, I personally think it's good for some images and bad for others, so I just output both and pick whichever turns out best.

Here's an example image using the normal Wan FP32 VAE: https://ibb.co/fGtZwdW8

And now the same image using the upscale2x VAE: https://ibb.co/RTj2DjVw

It's not in the workflow by default. To use it, you need to grab the ComfyUI VAE Utils node set and use the "VAE Decode (VAE Utils)" node instead of the regular VAE decoder. Then you also need to downscale your image by 50%, because this VAE decodes the image at 2x resolution (which is why it's so sharp). This pic shows what the setup should look like: https://ibb.co/XcrXmpr

What About Non-Realistic Images?

I still recommend using the raw model with the turbo lora at 0.6 strength for this UNLESS you're using a non-realism turbo finetune. This is because the raw model is much more creative than the turbo model; you'll get better variety this way.

If you're doing non-realism using the RAW model you should generally switch the second sampler off because you usually don't need/want the extra detailing step. That's good because it saves gen time! Just remember to turn it back on if you're doing realism again ;)

Alternatively, if you want to get started with a turbo anime finetune I recommend See-Krea2 - you can use it with the full-turbo workflow I've provided, just don't use the filter bypass lora with it (it's unnecessary). It can do You can also use CFG = 2 with it if you want to use a negative prompt. The author says not to use CFG with it but idk why, it works great. Use euler/beta 8 steps , which is the default in the workflow when you download it.

Anime Tags

I recommend putting the following tags at the start of your prompt if you're doing anime, no matter which finetune or loras you use; it'll help stabilise the style. I tested these quite a lot & there's no wasted terminology. However, if your finetune has built-in styles or a very strong main style you may want to remove or replace the tags before "masterpiece", those are style tags. I use all these tags with the See-Krea2 model I recommended above, but it's capable of doing other styles so you should read its info page and adjust if you want.

clean anime style, digital anime, smooth anime, cinematic anime, smooth shading, masterpiece, perfect quality, highly detailed, newest, (score_9, score_8, score_7:0.25).

You can control things slightly further with the below tags as well, just slot them in before the "masterpiece" tag if you want them. There are probably more, these are just the ones I've tested extensively:

"cel-shading" -> makes things have a more flat & less detailed anime look, rather than detailed or 3D

"mature" -> makes the faces of characters look a bit more adult-ish & less cutesy, it has no other effects so you can safely just leave it in if you prefer it

"detailed illustration" -> more noisy/detailed/painting-like style, depending on the specific finetune & loras you're using

Full Turbo Workflow?

You'll lose the creativity of the raw model by using it, but that may not matter to you at all depending on what you're doing. Or maybe you just need the speed. Or maybe you want to use a turbo finetune.

As mentioned earlier, the full turbo workflow is set up for making non-realistic images, like anime / concept art / digital paintings. It only has one sampler because you don't need an additional detailing step, and you don't really need the benefits of an ultra high-noise schedule either.

euler/beta is my general go-to for non-realistic images, and it holds up extremely well for Krea 2 in my testing so far. It also works for the base turbo model when doing realism. However, a lot of the realism finetunes work better with euler/simple instead, so always check the recommended settings when you download one.

If you're not using a finetune, you can either use the raw main model + turbo lora at 1.0 strength, or just use the turbo main model without the turbo lora. Using a turbo main model without the lora saves you about ~1.5GB of RAM (regular RAM, not VRAM), so that can be handy if you have low system RAM. No effect on VRAM so don't worry about that.

Krea 2's Facial Expression Problem: Censorship

This is a big problem, and as of 2026-09-18 it's still not fully resolved.

Basically, you'll see discussions going around about how Krea 2 doesn't do a very good job with facial expressions; characters lack expressiveness, and seem to have "dead eyes" a lot of the time. Smiles don't reach the eyes, people refuse to make any expressions even when you try to prompt for it, that sort of thing. It's nearly impossible to make someone look angry, fierce, or anything more than mildly annoyed , and the model will flat out refuse to make anyone smile properly at all.

This is a very common problem with distilled models (i.e. turbo models), but in Krea's case it's mostly because of ridiculous censorship they baked into the model. The developers heavily censored Krea 2 against whatever content they arbitrarily decided was 'harmful', and in doing so they lobotomised their own model, because censoring a model in post-training always damages the model. Krea 2 knows how to make an angry face, it just won't do it because it was collateral damage during the lobotomy.

You literally can't make people smile properly in most images with Krea 2 due to the censorship. That's not an exaggeration, try generating someone with a natural, realistic smile. Dumbest thing I've seen in years.

Luckily you can partially bypass the censorship using a simple lora, which you should use even if you're doing SFW stuff . It just makes better images, period. Some people say it also reduces the detail in the images, which is true - but this is actually just because you need to cook them a little longer. That is to say, if you have good sampler settings it's no problem. But it only works up to a point.

This workflow recommends using the bypass lora at 1.0 strength, but sometimes you need to go higher - even for SFW prompts - to get what you need. This isn't good because it degrades the image quality, but that's censorship for you. Stick to strength 1 if you can. We'll need finetunes to properly decensor the model (as of 2026-09-18 I haven't found any good RAW finetunes yet, but there are good...

Source context: 4023 downloads · Type Workflows · Base model Krea 2

VAE · Hugging Face · Kijai/WanVideo_comfy · Kijai/WanVideo_comfy · Hugging Face