My advice, use:
V2 & V3 for graphic design, illustration, and painting
V1 & V4 for photorealism
V4 (bf16 & Int8 ConvRot)
There's no point in chasing two hares at once, Variant 4 is oriented exclusively towards the photorealism, The pictorial aspect will be considered exclusively in subsequent versions..
V4 has been completely uncensored, if any censorship persists, it stems from the Qwen3VL text encoder, though this remains very rare.
The colors have been kept to a very neutral and natural palette.
V3 (bf16 & INT8 ConvRot) :
Improved skin texture for greater realism
Optimized lighting effects, more subtle and with better contrast
Enhanced painterly visual style
Corrected color shift toward warm tones, color rendering is more neutral and realistic
Feedback on V3
V3 is a lot more restrictive
Yes, the "unlocked" feel has been diluted; I’ll need to revisit that in future versions.
Colors are less realistic.
Yes, that aspect still needs fine-tuning in subsequent versions.
Less realistic than V1
That’s possible, given that I’m aiming to finalize a versatile model, one that is realistic for photography yet also graphic in style.
KREA 2 TURBO Vs KREAMANIA
V2 (bf16 & fp8) : Improvement of version 1 by merging an unpublished personal LoRAs to optimize lighting, textures, and graphic effects.
V1 (fp8) : The removal of certain restrictions compared to the model turbo fp8 scaled, along with improved realism and nudity rendering. The goal wasn't to focus on an NSFW model (even though it is capable of that), but rather to unlock the full range of creative possibilities.
Recommended ComfyUI extensions:
https://github.com/ethanfel/ComfyUI-Krea2TextEncoder
Description
Steps : 8
Sampler : er_sde
Scheduler : Simple
FAQ
Comments (10)
Thank you for version 4, I will definitely try it out.
you had lightning in a bottle with v1. Would love a convrot version.
Just tried V4 and it's not as good for me. It's higher fidelity and better skin texture, but in a studio photography hyper realistic overcooked way and doesn't understand concepts as well as V1.
Yes, I see your point. In practice, if you try to improve one aspect of the model, such as skin rendering, you inevitably lose out elsewhere. Achieving realistic texture essentially requires studio-shot datasets, which impacts the model's overall lighting. Conversely, opting for less detail means relying on outdoor photos without additional lighting, resulting in more authentic illumination. Making these choices inevitably entails compromises.
I will publish a V1 ConvRot version in the coming days.
Hello, in my opinion, this version has everything needed. It generates art excellently, and the realism, in my view, is quite good. It doesn't have the excessive sharpness present in version 3. I think that if you refine the prompt itself and install the ComfyUI-Krea2T-Enhancer node, prompt adherence will improve. The reason I didn't like version 3 much was that it handled styles very poorly. Version 3 is really a separate, unique model with its own distinct character.
Hi Zak, I’m really glad to see that the versions are sparking a debate. We all have different expectations, which is why our assessments differ. Personally, V3 is my favorite; it allows my LoRAs to fully express themselves—and does so in a truly masterful way compared to the other versions. Yet, V3 is the least appreciated and most criticized one.
I feel there is more to explore in the pictorial style than in realism, hence my interest.
@Adel_AI I agree, and I'm grateful that you released version 4. I really like your models. Perhaps in the future I'll give version 3 a try for art generation. I just rarely use LoRA; I strive to get everything from the model itself and what you've built into it. I only use LoRA when I want something new. Thank you once again for your hard work.
V4 is still plastic. You should change training data to only real high quality photos.
Looking at my showcase, you can clearly see that the result is far removed from a "plastic" look. The issue stems from your prompt, which fails to leverage the model's capabilities, rather than the dataset, which is based on high-quality photos.
@hardcore568761 Craft the prompt correctly, and you'll be happy. This model handles human skin rendering excellently.
Any chance we can get v1 int8 convrot please? I really like its subtle effect but fp8 is 1\3 slower for me.