AI Influencer Master Class - Krea 2 Image-to-Image Uncensored [VERSION 3 OUT NOW] (Patreon)

Give this workflow a reference photo and it recreates the scene with your character in it, running on the new uncensored Krea2 V2.5 Turbo model. It works differently from my SFW image-to-image. Instead of painting over your photo, it has the AI look at your photo and write a detailed description of it (the pose, the act, the setting, the lighting, what's being worn), puts your trigger word in front, and then generates a brand new image from that description. You get the same scene with your character, not a pixel copy of the original.

Your result can only be as good as your reference. A sharp, well lit photo gives you a great result. (The reference photo is blurred in these screenshots on purpose.)

What's new in version 3

A new model built for this. Both uncensored workflows now run on a new Krea2 V2.5 Turbo model trained for uncensored content, instead of the base Krea 2 Turbo model, so anatomy and poses come out a lot more reliable. The download link is at the bottom.

The hair texture problem is fixed. A lot of you saw a crunchy, almost painted texture in hair. I traced it back to the regular KSampler, so this version swaps it for the ClownsharKSampler from RES4LYF. Same speed, clean hair.

New text encoder. All my Krea 2 workflows now use the same one, qwen3-vl-4b-heretic_int8_dynamic_convrot.

One model does the captions too. The new text encoder also writes the description of your reference photo, so you no longer need a separate vision model.

Cleaner layout. Everything sits in labeled boxes that read left to right, in the order the image actually gets made. The How To Use note on the far left has all the download links too.

Optional film grain. There's a film grain node at the end. It's off by default.

The face detailer is gone. It was giving bad results with Krea 2, so only the eye pass is left.

How to use it

1. Load your reference photo

In the Reference Image box, drag your photo onto the LOAD REF IMAGE node, or click choose file to upload. Pick one where the pose and what's happening are easy to see. The AI can only describe what it can clearly make out.

2. Load your character LoRA

In the Models box, find the Power Lora Loader. The last slot is your character. Right now it has my Seraphina LoRA in it, so click the name and pick your own. Start with the strength at 1.0.

The slots above your character are the uncensored, realism and detail LoRAs I tuned for this workflow. Each one has a small on/off toggle on its left. Turn off the ones your scene doesn't need. For example, switch the oral LoRA off when the scene isn't oral, otherwise it keeps pulling your images in that direction. Download links for all of them are at the bottom.

3. Put your trigger word in

There's no prompt box to write in here, because the AI writes the scene for you. The only thing you add is who should be in it. In the CR Prompt List node, the first box (prepend_text) says s3ra, a woman,. Replace s3ra with your LoRA's trigger word, and keep the comma and the space at the end.

The last box adds a line about a candid, iPhone style photo to the end of every prompt. Leave it as it is. The AI's description of your photo goes in between the two automatically.

4. The auto caption, and where to read it

The Generate Text node is the part that looks at your reference photo and writes the description. The instructions it follows are typed into that node. They tell it to describe the scene in detail, and to never describe the woman's face, hair or body, so your character LoRA stays in charge of how she looks.

Below it, the Preview as Text node shows you exactly what the AI wrote on each run. If a result comes out wrong, read this first. Most of the time the description missed something or got something wrong, and that tells you what to fix. You can edit the instructions in Generate Text if you want a different style of description.

5. Set your image size

The width and height in the Empty Latent Image node set the size of your image. The default is 720 x 1280. Match the shape of your reference photo, so a landscape photo gets a landscape size, and the recreation comes out closer to the original.

6. The sampler (ClownsharKSampler)

These are tuned for Krea 2 Turbo, so leave them alone unless you're testing something on purpose. For reference, they're set to eta 0.5, sampler linear/ralston_2s, scheduler simple, 8 steps, cfg 1 and bongmath on.

You'll notice the seed says 42 and fixed. That's on purpose. The Smart Seed Variance node (a couple of steps down) changes on every run instead, and that's what gives you a different image each time you hit Queue.

The denoise here stays at 1.0. That's right for this workflow, because it's making a new image from the description rather than editing your photo.

7. Shift

At the top of the Prompt box, the ModelSamplingAuraFlow node sets the shift to 6. Shift decides how the model splits its steps between building the overall picture (the pose, the layout, where things go) and finishing the fine detail. The sampler settings above were tuned together with 6, so keep it there.

8. Smart Seed Variance

This node is the reason your images don't all come out looking the same. On every run it nudges your prompt a little, so you get changes in pose, framing and composition even when the prompt stays exactly the same.

If your images all look too similar, raise fine_tune_variance (it starts at 50).

If your character's face starts drifting and stops looking like her, lower it.

Everything else in this node can stay as it is.

9. Eye detail

Once the image is made, the EyeDetailer finds the eyes, zooms in and redraws just that small area with more detail. Small eyes in a full body shot usually come out mushy, and this pass is what fixes that.

The denoise (set to 0.10) is how much it's allowed to change the eyes. Nudge it up toward 0.15 or 0.2 if you want sharper eyes. Past about 0.25 it starts redrawing the eyes instead of cleaning them up, and your character can stop looking like herself.

This step needs the Eyeful and SAM models from the install section below. Without them it throws an error.

10. Film grain (optional)

Right before the image saves, there's an Image Film Grain node. It's switched off (bypassed) by default, which is why it looks faded. To turn it on, click the node and press Ctrl+B. Press Ctrl+B again to switch it back off. It adds a light layer of grain so the final image looks more like it came off a real camera and less like a perfectly clean render.

Your finished images save into your ComfyUI output folder. You can change the file name and subfolder in the Save Image node.

Installing the required models

Every model here is a one time download. Put each file in the folder shown, then restart ComfyUI (or press R in ComfyUI to refresh the model lists) and it will show up in the dropdowns. Your models folder lives at ComfyUI/models/ inside your ComfyUI install.

Inside models/loras/, make two new folders named KREA2MISC and KREA2CHARS. The LoRA loader looks inside those two folders, so once your LoRAs are in there they all load automatically. If you'd rather keep your LoRAs loose in models/loras/, that works too. You'll just need to click each LoRA slot once and pick the file again.

Civitai needs you to be logged in to download, and for the uncensored models you also need mature content turned on in your Civitai settings, or the pages won't show the download button.

The main models

Diffusion model: the new uncensored Krea2 V2.5 Turbo model goes in models/diffusion_models/. Civitai page · Direct download (fp8 file). The page has four versions of the file. Grab the fp8 one (about 13 GB). Once it's downloaded, pick it in the Load Diffusion Model node.

Text encoder: qwen3-vl-4b-heretic_int8_dynamic_convrot.safetensors goes in models/text_encoders/. Direct download. Make sure you grab this exact file. There are a few similar looking Qwen3-VL heretic files floating around, and they are not the same.

VAE: qwen_image_vae.safetensors goes in models/vae/. Direct download

The LoRAs (all go in models/loras/KREA2MISC/)

The number after each name is the strength I have it set to in the workflow.

Famegrid-Natural-V1-Krea-2.safetensors at 0.55. Natural smartphone look (Famegrid Krea 2 Natural V1). Download on Civitai

The Krea 2 uncensored LoRA (version v4.3_EXP) at 0.5. Unlocks uncensored content in Krea 2. It downloads as KNP_000003000.safetensors. Either rename it to match the name in the second LoRA slot of the workflow, or just click that slot and pick the file. Download on Civitai

fedor_bypass.safetensors at 1.0. Krea 2 filter bypass. The file is tiny (about 1 KB), that's normal. Download on Civitai

Amateur_Nudes_v1_Krea2.safetensors at 0.5. Amateur, real-photo nude look (the Krea2 version). Download on Civitai

blowjob_krea2_v1.safetensors at 0.5. Oral scenes. Turn it off when your scene isn't one. Download on Civitai

Detailer-KREA2.safetensors at 1.5. Overall detail slider. Download on Civitai

Krea2Pussy.safetensors at 0.5. Anatomy detail. Download on Civitai

Your character LoRA at 1.0. Whatever you trained, in models/loras/KREA2CHARS/.

The eye detailer models

Eyeful_v2-Paired.pt goes in models/ultralytics/bbox/. Direct download

sam_vit_b_01ec64.pth goes in models/sams/. Direct download (Meta)

Custom nodes

Open ComfyUI Manager, click Install Custom Nodes, search each name and install it. Or open the workflow and use Install Missing Custom Nodes, which finds them for you.

RES4LYF (the ClownsharKSampler). GitHub

rgthree-comfy (the Power Lora Loader). GitHub

ComfyUI Impact Pack (the eye detailer). GitHub

ComfyUI Impact Subpack (the eye finder). GitHub

RBG Smart Seed Variance. GitHub

WAS Node Suite (the film grain). GitHub

Comfyroll Custom Nodes (the CR Prompt List). GitHub

Install commands (Windows)

Open your ComfyUI folder in File Explorer (the folder that has the models folder inside it). Click the address bar at the top, type cmd and press Enter. That opens a Command Prompt already sitting in the right place. Paste these in:

mkdir models\ultralytics\bbox mkdir models\loras\KREA2MISC mkdir models\loras\KREA2CHARS curl.exe -L -o models\text_encoders\qwen3-vl-4b-heretic_int8_dynamic_convrot.safetensors https://huggingface.co/ethanfel/Qwen3-VL-4B-Heretic-ComfyUI-ConvRot-INT8/resolve/main/qwen3-vl-4b-heretic_int8_dynamic_convrot.safetensors curl.exe -L -o models\vae\qwen_image_vae.safetensors https://huggingface.co/Comfy-Org/Krea-2/resolve/main/vae/qwen_image_vae.safetensors curl.exe -L -o models\ultralytics\bbox\Eyeful_v2-Paired.pt https://huggingface.co/Tenofas/ComfyUI/resolve/main/ultralytics/bbox/Eyeful_v2-Paired.pt curl.exe -L -o models\sams\sam_vit_b_01ec64.pth https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth

If a mkdir line says the folder already exists, that's fine, ignore it. Type curl.exe exactly like that, with the .exe. In PowerShell, plain curl runs a different command and these lines fail. The diffusion model and the LoRAs aren't in here because Civitai needs a login, so download those from the links above in your browser.

Install commands (RunPod / JupyterLab)

On RunPod your ComfyUI lives at /workspace/ComfyUI, so you can paste these into the JupyterLab terminal from anywhere:

mkdir -p /workspace/ComfyUI/models/ultralytics/bbox /workspace/ComfyUI/models/sams /workspace/ComfyUI/models/loras/KREA2MISC /workspace/ComfyUI/models/loras/KREA2CHARS curl -L -o /workspace/ComfyUI/models/text_encoders/qwen3-vl-4b-heretic_int8_dynamic_convrot.safetensors https://huggingface.co/ethanfel/Qwen3-VL-4B-Heretic-ComfyUI-ConvRot-INT8/resolve/main/qwen3-vl-4b-heretic_int8_dynamic_convrot.safetensors curl -L -o /workspace/ComfyUI/models/vae/qwen_image_vae.safetensors https://huggingface.co/Comfy-Org/Krea-2/resolve/main/vae/qwen_image_vae.safetensors curl -L -o /workspace/ComfyUI/models/ultralytics/bbox/Eyeful_v2-Paired.pt https://huggingface.co/Tenofas/ComfyUI/resolve/main/ultralytics/bbox/Eyeful_v2-Paired.pt curl -L -o /workspace/ComfyUI/models/sams/sam_vit_b_01ec64.pth https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth

For the diffusion model and the LoRAs on RunPod, download them on your computer and drag them into the right folders in JupyterLab's file browser.

That's everything. No checkpoint or upscale model needed for this one.

Workflow by LoRAtech