run-comfyui-image¶
Release Notes Policy¶
- A new dated subtitle is added whenever a new stable ComfyUI release is included.
- The date is updated later when custom nodes are added, removed, or changed, or when code improvements and bug fixes are applied.
- The date reflects when the related custom nodes were cloned from their upstream repositories.
- Pinned custom nodes are not updated automatically, because changing those pins can introduce compatibility issues.
- Tested on a wide range of GPU's including RTX 3090,4090,5090, L40S, L4 , PRO 6000 MiG 24/48 Gb.
14102026¶
- Updated ComfyUI to version 0.40.0.
- ComfyUI_Nvidia_VFX_Nodes as shown on ComfyUI
13102026¶
- Updated ComfyUI to version 0.39.0.
- ComfyUI-qwen_img_2_1_enhancer
- ComfyUI-Omnichar
- Updated QwenImageRefPack with an extra omnichar load and edit node.
- Updated Omnichar adding uploading and downloading char files, extending slots to 9 images.
- Updated prompt enhancer Qwen Image 2.1 to use multiple models.
- Updated workflows Qwen Image 2.1 for turbo lora and phrase weight.
- ComfyUI-EditUtils
- ComfyUI-qwenmultiangle updated for Qwen Image 2.1
- comfyui_controlnet_aux
- comfyui-poser
06102026 image¶
- ComfyUI-Omnichar
- Updated QwenImageRefPack with an extra omnichar load and edit node.
- Updated Omnichar adding uploading and downloading char files, extending slots to 9 images.
02102026¶
- ComfyUI 0.38.0
- Fixed Krea-2 prompt enhancer.
01102026 image¶
- Updated customs nodes.
- Updated Flux-Klein SAM3 workflows.
- Lowered volume of pod.
- Updated qwen edit workflows.
29092026¶
- ComfyUI 0.37.1
- Pytorch 2.12.1 CUDA 13.0 build
- Added Runpod templates with workflows for QWEN Image 2.1 with/without heretic prompt enhancer.
- ComfyUI-QwenImageRefPack
- Updated llama-cpp to
b11115 - Updated Krea-2 workflows.
Container CUDA upgrade notice¶
- run-comfyui-image2 requires CUDA 13.x (due to breaking changes in "comfy-kitchen" and no support for CUDA 12.x on modern GPU's by ComfyUI).
- If you need due GPU shortage a lower cuda version change runpod container template tag to run-comfyui-image2:18092026.
- run-comfyui-image will be pinned on ComfyUI 0.36.0, custom nodes with still be maintained.
18092026 image and image2¶
- ComfyUI 0.36.0
- Removed unused custom nodes.
- Forked and customized model_linker.
- Updated Lora-Manager template.
- Lowered pod size by excluding unused media in custom_nodes.
- Last version compatible with cuda 12.x and higher.
12092026¶
- ComfyUI 0.35.0
- Added SenseNova U1.5 to image2
- Updated image
08092026¶
- ComfyUI 0.34.0
- Added native llama-cpp to image2 with CUDA support
- comfyui-obvpm
- ComfyUI-CyberKrea-Sampler
12082026 image2¶
- ComfyUI 0.32.0
11082026¶
- ComfyUI 0.31.0
- krea2-anypaint
- Krea-Moodboard
- Krea2-NAG
Split¶
Krea-2 has been moved to image2 to reduce the number of custom nodes in image and ensure compatibility with future models. Both images will continue to be updated as long as ComfyUI supports their respective PyTorch/Cuda versions.
05082026 image¶
- ComfyUI 0.30.0
- comfyui-model-linker
01082026¶
- ComfyUI 0.29.0
- Fixed logic error with hf download watchdog.
- Updated deployment tutorial.
20072026¶
- ComfyUI 0.28.0
- Fixed Flux-Klein text_encoder download.
- Updated hf download logic trying to prevent xet stalls.
- one-node-flux-2-klein
- Krea2edit
- krea2-negpip
- Krea2-Projector-Tuner
- Krea2-attention-tweak.git
- Krea2-Ostris-Edit
09072026¶
- ComfyUI 0.27.0
- krea-reference
- KreaReason
- Krea2-StyleTransfer
30062026¶
- ComfyUI 0.26.0
- Added Krea-2 t2v uncensored workflow.
- krea2-conditioning
- Krea2TextEncoder
- Conditioning-Rebalance
- Krea2T-Enhancer
19062026¶
- ComfyUI 0.25.0
- KleinTiledUpscaler
- JoyCaption
- Fixed typos, white space breaks, documentation with Codex.
- Made moves and hf downloads more verbose.
- Fixed Flux-Klein download.
06062026¶
- ComfyUI 0.24.0
- Untwisting-RoPE
- TextEncodeEditAdvanced
- Removed VLM as prompt enhance works better with a separate pod (textgen for example).
- Removed ComfyUI-SecNodes as SAM3 is more powerful.
01062026¶
- ComfyUI 0.23.0
- Flux2KontextConditioner
25052026¶
- ComfyUI 0.22.0
- KleinRefGrid
- ReferenceLatentPlus
- Inpaint-CropAndStitch
15052026¶
- ComfyUI 0.21.1
- Easy-Use
- S2Guidance
09052026¶
- ComfyUI 0.20.1
- Updated Civitai downloader script and environment variables
- NKD-Klein-Tools
27042026¶
- ComfyUI 0.19.3
- DiversityBoost
- Flux2Klein-Conditioning-Toolkit
- Iamccs-nodes
16042026¶
- ComfyUI 0.19.1
- ERNIE image support
15042026¶
- ComfyUI 0.19.0
- Updated ZIB-ZIT workflows, QWEN workflow.
- Z-Image CLIP text encoder changed to abliturated
- Lora Optimizer
24032026¶
- ComfyUI 0.18.1
- QwenMultiAngle
06032026¶
- ComfyUI 0.16.3
26022026¶
- ComfyUI 0.15.0
- EasyLoRAMerger
22022026¶
- ComfyUI 0.14.0
- QwenVL-mod
- ZImageTurboHQNodes
- QWEN image controlnet.
11022026¶
- ComfyUI 0.13.0
- CameraAngleSelector
04022026¶
01022026¶
- ComfyUI 0.11.0
- Updated custom_nodes for Flux-Klein edit, updated QWEN EDIT template and workflows
- Z-Image Base.
25012026¶
- ComfyUI 0.10.0
- Added configuration for loading workflows and models depending on the available VRAM.
- Arthemy_Live-Tuner-ZIT-ComfyUI
- Flux2ImageReference
- ZImagePowerNodes
- Flux2Klein-Enhancer
20012026¶
- ComfyUI 0.9.2
14012026¶
- ComfyUI 0.9.1
- Added environment variable MEDIA[1-50] to download files to /workspace/ComfyUI/input for test/tutorial purpose.
11012026¶
- ComfyUI 0.8.0
- Added environment variable "CIVITAI_MODEL_UNET_URL" for download.
- Fixed civitai downloader reloading models after restarting pod.