JLC Flux2 ControlNet v1.0.0: FLUX.2 ControlNet for ComfyUI

news

Damkohler released JLC Flux2 ControlNet v1.0.0, a custom node bringing FLUX.2-dev Fun ControlNet Union into ComfyUI with multi-branch control, reference conditioning, and in/out-paint.

JLC Flux2 ControlNet v1.0.0 is a custom node pack by Damkohler that integrates the FLUX.2-dev Fun ControlNet Union side model into ComfyUI. It adds multi-ControlNet composition, multi-reference conditioning, reusable latent caching, and experimental in/out-paint support — all without replacing ComfyUI's core FLUX.2 pipeline.

Installation

Clone into ComfyUI/custom_nodes/:

cd ComfyUI/custom_nodes
git clone https://github.com/Damkohler/JLC-Flux2-ControlNet.git

You also need the FLUX.2-dev Fun ControlNet Union checkpoint in ComfyUI/models/controlnet/:

Download FLUX.2-dev Fun ControlNet Union (2.38 GB)

Requirements

  • ComfyUI with FLUX.2 support
  • FLUX.2-dev diffusion model + text encoder + VAE
  • Python 3.10+

Included Workflows

The repository ships with three ready-to-use workflows. Drag the PNG into ComfyUI to load, or use the JSON file:

1. Basic Reference-Image & Multi-ControlNet

Basic workflow

2. Experimental Inpainting

Inpainting workflow

3. Full Reference, Multi-ControlNet, Inpainting & Cache

Full workflow

Node Reference

NodeStatusPurpose
JLC Flux2 ControlNet LoaderStableLoads the compact side model as a reusable ControlNet object
JLC Flux2 ControlNet ApplyStableAttaches one ControlNet branch to conditioning
JLC Flux2 ControlNet Apply AdvancedStableShared config for positive and negative conditioning
JLC Flux2 ControlNet OrchestratorStableFlat 1-4 branch composition sharing a single loaded model
JLC Flux2 Reference Image OrchestratorStableUp to 10 reference images with per-slot enable controls
JLC Flux2 ControlNet Latents CacheStableCPU cache for up to 4 active control images
JLC Flux2 In/Out-Paint AdapterExperimentalMask-aware inpainting/outpainting adapter
JLC Flux2 Inpaint Context CacheExperimentalPrecomputed inpaint context caching

Notes

  • The project does not include image preprocessors (pose, depth, edge, etc.) — use ComfyUI's built-in nodes or the optional companion pack JLC ComfyUI Nodes
  • The FLUX.2-dev Fun ControlNet Union side model is required and can be downloaded from alibaba-pai/FLUX.2-dev-Fun-Controlnet-Union
  • Licensed under MIT
JLC Flux2 ControlNet on GitHub

Comments

Sign in with GitHub to join the discussion.

Loading comments…
JLC Flux2 ControlNet v1.0.0: FLUX.2 ControlNet for ComfyUI | ComfyUI Wiki