Join The Tinkerer on Whop. Membership gets you early releases, private tools and a bunch of extra stuff.
👉 Join on Whop
💬 Join the community for support, free tools and early news on Discord
What's New in V4.0
CyberRealistic Krea 2 Workflow V4.0 — A More Flexible, Modular Workflow
Version 4.0 is a significant update to the CyberRealistic Krea 2 workflow. The proven generation pipeline remains, but the workflow has been reorganized around a collection of independent, optional tools.
The previous Cyberdelia Prompt Controls / Prompt Engineer integration has been removed. Instead of relying on a single prompt-processing system, V4.0 gives you direct control over your prompts, camera styles, character composition, model enhancements, and image metadata.
The idea is simple: use what you need, disable what you don't.
What's New
Camera Prompt Presets
The new Cyberdelia Camera Prompt Presets node makes it easy to give images a particular photographic character without rewriting your entire prompt.
Choose from predefined camera and photographic-process presets, including:
Early mobile-phone and compact digital camera photography
Analog film and historical photographic processes
Professional studio and editorial photography
Lens effects and optical characteristics
Vintage, imperfect, and everyday snapshot aesthetics
Presets are organized into JSON libraries, and you can create or import your own collections. The module has its own ON/OFF switch, so you can return to your original prompt at any time.
Two Character Composer
The Krea2 Two Character Composer introduces one of the biggest additions to V4.0: the ability to work with two different characters, each with their own regional LoRA, within a single image.
You can:
Assign a separate LoRA, trigger word, and prompt to each character.
Divide the composition left/right or top/bottom.
Position characters using adjustable regions, free boxes, or custom masks.
Control the separation between characters.
Keep the shared scene, lighting, and background consistent.
This is particularly useful for scenes featuring two distinct people whose identities need to remain separate. The Composer is completely optional. With its switch set to OFF, the workflow follows the normal generation path.
CyberKrea Enhancer Advanced
The CyberKrea Enhancer Advanced provides optional controls for Krea 2's text-fusion processing. It allows experimentation with how the model handles text information during generation, including separate adjustment of text-token scaling. The enhancer is disabled by default and can be enabled when you want to experiment with its effect on your results.
Negative Prompt Support with NegPiP
The Cyberdelia NegPiP implementation provides negative-prompt support for Krea 2, including workflows running at CFG 1. This allows you to enter unwanted characteristics or visual elements in a dedicated negative prompt without switching to a conventional high-CFG generation setup. Negative prompting is integrated into the main workflow and is also supported by the Two Character Composer.
Wildcard Prompts
V4.0 includes an optional Wildcard Prompts module for working with lists of prompts and randomized selections. You can prepare multiple scene descriptions, select a random prompt for each generation, or work with an existing prompt collection. A dedicated ON/OFF switch lets you choose between your normal prompt and the wildcard-generated prompt. This is useful for testing different ideas, generating varied scenes, or running larger batches without manually changing your prompt every time.
Improved Image Metadata
The workflow uses Cyberdelia Image Metadata to save generation information directly with your images.
It supports Civitai-compatible metadata, including:
Prompts and negative prompts
Model information
LoRAs and their strengths
Seeds and sampling settings
CyberKrea sampler parameters
Two Character Composer information when enabled
The metadata system is designed to handle complex workflows with multiple samplers, optional modules, and switchable conditioning paths. This helps preserve the settings used to create your images.
The Generation Pipeline
V4.0 continues to use the CyberKrea Sampler as its main generation engine.
The familiar workflow structure remains available:
Model, text encoder, and VAE loading
Resolution and seed controls
Optional LoRA loading
First-pass generation with CyberKrea Sampler
Optional latent upscaling and second-pass refinement
Optional Face Detailer
Image previews and before/after comparisons
Final image saving with metadata
The optional modules can be enabled or disabled independently, allowing you to keep the workflow relatively simple or use more advanced features when needed.
Prefer AI-Generated Prompts? Try Scene Director
If you previously used Prompt Engineer to create or expand prompts with a local language model, take a look at Scene Director. Scene Director is a separate application for creating varied, detailed image prompts using a local LLM, such as one running through LM Studio. Rather than simply rewriting an existing prompt, it can develop different scenes, situations, compositions, and visual ideas from a short description. It supports both Krea 2 and Z-Image Turbo, and can send generated prompts directly to a configured ComfyUI workflow. Scene Director is actively being developed and will be expanded with additional features and improvements over the coming period. It is entirely optional. V4.0 does not require an LLM, LM Studio, or Scene Director to generate images.
Installation and Plugins
The workflow uses several custom nodes developed or maintained under the Cyberdelia project.
You can find the individual repositories and installation instructions here:
Cyberdelia's ComfyUI plugins can also be found through the Cyberdelia publisher page on the Comfy Registry, where available releases can be installed through compatible ComfyUI management tools. Additional dependencies, including the CyberKrea Sampler, wildcard processing, rgthree, and optional detailing/upscaling nodes, may be required depending on which workflow modules you use. Make sure the necessary custom nodes are installed and updated before loading the workflow.
Final Notes
V4.0 is not intended to make prompt writing more complicated. Quite the opposite. The workflow now separates prompt creation, photographic styling, regional character control, sampling, and post-processing into individual components. You can simply enter a prompt and generate an image, or enable the additional tools when a scene requires more control.
More flexibility, fewer dependencies on a single prompt system, and a workflow that can continue to evolve as new Cyberdelia tools become available.
What’s New in V3.5
Version 3.5 keeps the proven generation pipeline from V3.0, but introduces the new Cyberdelia Prompt Controls — Presets node.
This preset-aware control node brings the main prompt options together in one place: enter your prompt, choose between engineered (LLM) and passthrough (RAW), enable or disable Vision, and select the system-prompt preset used by Prompt Engineer.
The preset selector automatically follows the selected input type:
With Vision disabled, it shows the normal prompt-rewriting presets.
With Vision enabled, it switches to the dedicated image-to-prompt presets.
The complete selected system prompt is stored inside the workflow, keeping shared workflows reproducible even if an external preset is later changed.
Passthrough mode continues to work without LM Studio or another local LLM.
This makes the workflow a natural companion to the new Krea 2 prompt-preset library. You can use presets to turn the same starting idea into photographic realism, painterly artwork, editorial illustration, graphic-novel artwork, studio portraits, and many other targeted styles.
🆓 Preset guide and free starter selection:
📦 Complete Krea 2 and Z-Image Turbo preset collections (for Whop members):
https://whop.com/cyberdelia-ai-lab/exp_Q7Jc5e2sjBWTae/app/
A clean and practical CyberRealistic Krea 2 workflow for ComfyUI, built around the CyberKrea Sampler and designed to keep the main generation process simple while still offering useful extras such as Face Detailer, latent upscaling, LoRA support, prompt engineering, comparison previews, and improved saved-image metadata.
CyberRealistic Krea 2 - ComfyUI Workflow V3
The workflow keeps the parts you use most close together: model settings, prompt input, generation, previews, and saving. Optional modules stay out of the way until you need them.
Face Detailer and Upscaler can be enabled or disabled from the module toggle panel. Both are disabled by default, so the workflow starts with a straightforward Krea 2 generation.
What's included in V3
CyberKrea Sampler. The main generation pass uses the CyberKrea Sampler with Fast, Balanced, and Quality presets. Balanced is the default, while the more advanced sampler controls remain available when you want to fine-tune the result.
Krea 2 resolution presets. The CyberKrea Empty Latent node provides practical aspect ratios and resolutions without having to enter dimensions manually.
NegPiP prompt encoding. Positive and negative prompting is handled through the dedicated NegPiP node, including support for negative prompt weighting while keeping the Krea 2 conditioning pipeline intact.
Optional Prompt Engineer. Rewrite simple ideas or existing prompts into natural Krea 2 prompts using a local OpenAI-compatible LLM. Vision mode can also analyze a reference image and turn it into a Krea 2 prompt. Normal generation does not require a local LLM; keep the Prompt Engineer in passthrough (raw) mode to use your prompt directly.
LoRA loading. An rgthree Power LoRA Loader is included for quickly adding and combining LoRAs without rebuilding the model chain.
Krea 2 Turbo Enhancer. The optional Krea2T Enhancer is included in the model section and disabled by default.
Face Detailer. Optional YOLO face detection with SAM-assisted masking and Impact Pack Face Detailer. Useful for repairing or refining smaller faces without affecting the complete image.
SesquiLSR latent upscale. The optional Upscaler uses SesquiLSR to enlarge the latent, followed by a lightweight second Krea 2 sampling pass. No separate ESRGAN-style upscale model is required.
Comparison previews. Dedicated comparison nodes make it easy to check the original first pass against the Face Detailer and upscale results.
Improved image metadata. Final images are saved through Cyberdelia Save Image With Metadata for cleaner generation metadata and easier sharing.
Face Detailer and Upscaler are disabled by default. Enable only the modules you need and the final Save Image node follows the resulting image through the chain.
Prompt Engineer Settings
The Prompt Engineer is kept compact in the main workflow. To access its advanced settings, click the small open subgraph icon in the top-right corner of the 🧠 Prompt Engineer node.
Inside the subgraph you can configure the local LLM connection and generation settings, including API URL, model, temperature, max tokens, timeout, retries, output cleanup, and Vision settings.
The default API endpoint is:
http://localhost:1234/v1
This is intended for a local OpenAI-compatible server such as LM Studio. The model can normally remain set to auto.
You do not need a local LLM to use the workflow. With Prompt Engineer set to passthrough (raw), your prompt is sent directly to Krea 2 without calling the LLM.
For image-to-prompt generation, enable Vision and connect an image to the Prompt Engineer image input.
Requirements
A recent ComfyUI build with subgraph support
Krea 2 Model
qwen3vl_4b_bf16.safetensorsorqwen3vl_4b_fp8_scaled.safetensorsKrea 2 text encoderqwen_image_vae.safetensorsbbox/face_yolov8m.ptfor Face Detailer detectionsam_vit_b_01ec64.pthfor SAM-assisted Face Detailer maskingA local OpenAI-compatible LLM server only if you want to use the optional Prompt Engineer
Nodes / Custom Nodes Used
rgthree-comfy
https://github.com/rgthree/rgthree-comfyCyberKrea Sampler
https://github.com/cyberdeliaAI/ComfyUI-CyberKrea-SamplerCyberdelia Z-Engineer
https://github.com/cyberdeliaAI/comfyui-cyberdelia-z-engineerCyberdelia Metadata / Save Image With Metadata
https://github.com/cyberdeliaAI/comfyui-cyberdelia-metadataNegPiP for Z-Image / Krea 2
https://github.com/cyberdeliaAI/comfyui-negpip-zimageComfyUI Impact Pack
https://github.com/ltdrdata/ComfyUI-Impact-PackComfyUI Impact Subpack
https://github.com/ltdrdata/ComfyUI-Impact-SubpackKrea 2 Turbo Enhancer
https://github.com/capitan01R/ComfyUI-Krea2T-Enhancer
Prompt Engineer note: the workflow works normally with Prompt Engineer in passthrough mode. A local LLM is only required when you enable prompt rewriting or vision-based prompt generation.
Description
Join The Tinkerer on Whop. Membership gets you early releases, private tools and a bunch of extra stuff.
👉 Join on Whop