Nodes are now Loaded!
For now, install through GitHub if you want the most recent changes (i'm doing small hotfixes)
If you install the nodes through comfyui's manager Restart ComfyUI for the UI changes to load correctly
https://github.com/ApexLightSpeed/ComfyUI_ApexPipeline
V2 is a ***MAJOR*** rebuild of the original workflow. The generation process has been reorganized into four Apex nodes: Model Matrix, Generation Nexus, Archive, and Image Preview. The aim was to make the workflow easier to understand and use while keeping the familiar image-generation process intact.
New layout: Four Apex nodes arranged in one clear row.
Model Matrix: Model loading and runtime options gathered into one place.
Generation Nexus: Prompt, canvas, sampling, seed, and LoRA controls brought together.
Better LoRA controls: Separate model and CLIP strengths, −5 to +5 sliders, 0.05 step buttons, and fields for custom values.
Find triggers: Scans local LoRA metadata and can look up trigger words by file hash.
Archive: Saves images with workflow and generation metadata, plus a single adjustable last-save preview.
Image Preview: A separate, resizable node for viewing the generated image at a larger size.
Visual overhaul: New black-and-crimson Apex styling, original header artwork, and clearer sections and controls.
Cleaner starter workflow: Blank prompt and no preloaded LoRAs, so it opens ready for your own setup.
Description
First version!
Requirements
ComfyUI with Krea 2 support.
LC123 custom nodes.
Pixaroma custom nodes.
The required Krea 2 Turbo model, Qwen3-VL text encoder, and Qwen Image VAE.
Third-party dependencies and model files must be installed separately. No checkpoints or LoRAs are bundled with this workflow.