October 29, 2026
Summary: This update adds 19 custom nodes, with ComfyUI core upgraded to v0.43.0. Highlights: MiniMax-H3 Spectrum acceleration & prompt enhancement, MOSS-Image block cache, FitDiT virtual try-on, multi-stage sampler chains, high-quality image saving.Core Update
- ComfyUI version: v0.42.0 (
06b8412) → v0.43.0 (b8822fb) - Effective: immediately (refresh the page)
New Capabilities
New custom nodes added (19 total):- MiniMax-H3 / MOSS-Image Extensions
ComfyUI-Spectrum-MiniMax-H3(commit806fe49) — Spectrum acceleration for H3.comfyui-minimax-h3-prompt-enhancer-T8(commit6cf6610) — H3 prompt enhancer.Comfyui-Moss-Image-Blockcache-T8(commitbcb694c) — MOSS-Image block cache acceleration.
- Upscale / Super-Resolution
ComfyUI-VOSR2(commit6a21810) — VOSR 2.0 image super-resolution.ComfyUI-SeedVR2_VideoUpscaler-numz(commit4490bd1) — SeedVR2 v2.5.24 modular architecture.
- Face / Try-On
ComfyUI_FaceSimilarity(commit8bc7c55) — Facial similarity comparison.FitDiT-ComfyUI(commit17215e3) — FitDiT virtual try-on.
- Samplers
ComfyUI-TripleKSampler(commitb35509c) — Multi-stage sampler chains.tiled_ksampler(commitd954379) — Tiled KSampler for SDXL/SD1.5.
- Prompt / Captioning
joycaption_comfyui(commit8789c1c) — JoyCaption image captioning.ComfyUI-PascalEditor(commit478a9d9) — In-graph text editor v0.9.1.Comfyui_Redux_Advanced(commit38ebcbb) — Advanced Redux prompt processing.ComfyUI-Redux-Prompt(commit50ebbfd) — Apply image prompt/Redux style to current prompt.
- Detection / Video
ComfyUI-YOLO(commit750d8ae) — YOLO object detection.ComfyUI-PersonaLive(commit9ef67ed) — Persona Live nodes.
- Utility / Other
ComfyUI-Text_Image-Composite(commit47c1531) — Text and image composite.comfyui_lora_tag_loader(commitc1e1a06) — LoRA tag loader.ComfyUI-Olm-DragCrop(commit5e286e8) — Interactive drag/crop.ComfyUI-HQ-Image-Save(commit44b0c6f) — High-quality save (PNG/JPEG/WEBP/TIFF/EXR).
Bug Fixes
- Fixed
ComfyUI_roopload failure
Breaking Changes
Updated custom nodes:ComfyUI06b8412(v0.42.0) →b8822fb(v0.43.0)
- none
Usage Notes
- Restart the environment / refresh the page to apply
October 22, 2026
Summary: This update adds 27 custom nodes, with ComfyUI core upgraded to v0.42.0. Highlights: OpenPose editors, SAM3 segmentation, ControlNet-LLLite, H3-Continuum / IF_MemoAvatar extensions, multiple samplers & upscalers, face/matting/normal-lighting toolchain.Core Update
- ComfyUI version: v0.39.0 (
14e3c90) → v0.42.0 (06b8412) - Effective: immediately (refresh the page)
New Capabilities
New custom nodes added (27 total):- Pose / Segmentation / Detection
ComfyUI-openpose-editor(commit75ba64b) — OpenPose pose editor.ComfyUI-ultimate-openpose-editor(commita21cc3b) — Ultimate OpenPose editor with render.ComfyUI-TBG-SAM3(commit90e47e2) — Meta SAM3 segmentation nodes.ComfyUI-InstanceDiffusion(commit5f49181) — Multi-instance control generation.
- MiniMax-H3 Extensions
ComfyUI-H3-Continuum(commit7e14fd) — H3 Continuum extension nodes.ComfyUI-SuperUltimateVaceTools(commit93298fd) — Super Ultimate VACE toolset.
- ControlNet / LoRA / Samplers
ControlNet-LLLite-ComfyUI(commit103a4457) — ControlNet LLLite nodes.ComfyUI-LoRA-Optimizer(commit28ec4e7) — LoRA optimizer.ComfyUI-Extra-Samplers(commit52eac1b) — Extra samplers.ComfyUI-sampler-lcm-alternative(commit883d4ec) — Improved LCM alternative sampler.ComfyUI-FSampler(commit5479059) — FSampler custom sampler.FreeU_Advanced(commitcf3714d) — FreeU Advanced model patching.ComfyUI_FluxMod(commitc2dc1e2) — FluxMod model patching.comfyui_jankhidiffusion(commit8de02bc) — Jank HiDiffusion nodes.
- Upscale / Super-Resolution
ComfyUI_SuperScaler(commit4669aa0) — Super-resolution upscaler.SD-Latent-Upscaler(commit89543b0) — SD latent upscaler (small neural net).
- Image Editing / Matting / Face
ComfyUI-SDMatte(commit30575c6) — Interactive image matting (point/box/mask).comfyui-lama-remover(commit070c022) — LaMa inpainting remover.ComfyUI-See-through(commit98d754b) — See-through clothing removal.comfyui_facetools(commit59566c5) — Face extraction, paste-back, and variation tools.ComfyUI_MatAnyone_Kytra(commitac2c6e4) — Kytra’s MatAnyone implementation.
- Video Generation
Comfyui-PainterFLF2V(commitc81a68f) — Enhanced first-last-frame video motion nodes.ComfyUI-IF_MemoAvatar(commitbcc480c) — Memory-guided diffusion for expressive talking video.
- Utility / Other
MOSS-Image(commit19f0769) — MOSS image generation toolkit.ComfyUI-DeGrid(commit42ba59d) — Grid layout sampling.cg-image-filter(commit1602dbe) — Image filter nodes.save-image-extended-comfyui(commitd49387f) — Extended save (AVIF/WebP/JPEG + custom folders/filenames).comfyUI_TJ_NormalLighting(commitf02c609) — Virtual lighting based on normal maps.
Breaking Changes
Updated custom nodes:ComfyUI14e3c90(v0.39.0) →06b8412(v0.42.0)
- none
Usage Notes
- Restart the environment / refresh the page to apply
October 15, 2026
Summary: This update adds 16 custom nodes, with ComfyUI core upgraded to v0.39.0. Highlights: Qwen Image 2.1 multi-angle/block-cache nodes, MiniMax-H3 extension suite (6 nodes), SAM2 segmentation, ACE-Step music generation, AuraSR super-resolution.Core Update
- ComfyUI version: v0.37.0 (b5c8993) → v0.39.0 (14e3c90)
- Effective: immediately (refresh the page)
New Capabilities
New custom nodes added (17 total):- Qwen Image 2.1
- Comfyui-Qwen-Image-2.1-MultiAngle-T8 (commit d11a723) — Qwen Image 2.1 multi-angle control.
- Comfyui-Qwen-Image-2.1-BlockCache-T8 (commit 53c9d3f) — Qwen Image 2.1 block cache acceleration.
- MiniMax-H3 Extensions
- ComfyUI-Fizgig-H3-Tweaks (commit 5f8b48a) — H3 training-free tweaks.
- ComfyUI-MiniMaxH3_LatentUpscaler (commit 895e3c4) — H3 latent upscaler (upstream source).
- ComfyUI-H3-Latent-Relay (commit 16a2734) — H3 latent relay node.
- ComfyUI-H3-Latent-Upscaler-Mamad8 (commit e982377) — H3 latent upscaler.
- selflift-Avatar (commit dc8e545) — SelfLift Avatar variant for H3 video/audio masks.
- ComfyUI_Qwen_H3_Prompt (commit 1564f02) — Generate H3 prompts via local Qwen3.8-27B.
- Segmentation / Detection
- ComfyUI-SAM2 (commit 757034b) — Segment-Anything 2 segmentation.
- Image-Detection-Bypass-Utility (commit 5c787cf) — Image detection bypass (noise injection / FFT smoothing / pixel perturbation).
- Upscale / Super-Resolution
- ComfyUi_NNLatentUpscale (commit 08105da) — Latent-space upscaling.
- ComfyUI-AuraSR (commit 29c97cf) — AuraSR super-resolution.
- Music / Audio
- ComfyUI_ACE-Step (commit d4c2dc1) — ACE-Step music generation base model.
- Utility / Other
- Comfyui-In-Context-Lora-Utils (commit ede1f33) — In-Context LoRA utilities.
- ComfyUI-AusBoss (commit ec61f13) — Outpainting workflow nodes.
- images-grid-comfy-plugin (commit acad7c4) — Image grid preview.
- ComfyUI-Prompt-Studio (commit cac64a8) — Prompt studio.
Breaking Changes
Updated custom nodes:- ComfyUI b5c8993 (v0.37.0) → 14e3c90 (v0.39.0)
Usage Notes
- Restart the environment / refresh the page to apply
September 20, 2026
Summary: This update adds 20 custom nodes, updates 8, and removes 1. Highlights: Flux toolchain expansion (ControlNet / Enhancer / Relighting), virtual try-on & tiled generation, image tagging & story generation, and a full MiniMax-H3 suite upgrade.Core Update
- ComfyUI version: v0.34.0 (
d07ab68) → v0.37.0 (b5c8993) - Effective: immediately (refresh the page)
New Capabilities
New custom nodes added (20 total):- Flux Toolchain
MistoControlNet-Flux-dev(commit12931880) — Lineart/contour ControlNet for Flux1.dev.ComfyUI-FluxExt-MZ(commit00e4b70d) — MinusZone Flux1 utility node collection.ComfyUI-Flux2Klein-Enhancer(commit6804643b) — Flux.2Klein 9B enhancement node suite.ComfyUI-IC-Light-Native(commit40883a83) — Native IC-Light relighting implementation.
- Image Generation & Editing
ComfyUI-ZImagePowerNodes(commitb86ef467) — Z-Image / Z-Image Turbo node collection.ComfyUI_CatVTON_Wrapper(commite2ae9b62) — CatVTON virtual try-on.ComfyUI-TiledDiffusion(commita155b1ba) — Tiled Diffusion / MultiDiffusion + optimized VAE.ComfyUI-seamless-tiling(commit9225ed5d) — Seamless tiling, modifies Conv2d/PatchEmbed for tileable output.comfyui_HiDream-Sampler(commit98ad017c) — HiDream sampler wrapper node.
- Personalization / Portrait
ComfyUI-HyperLoRA(commit108d4c32) — Official HyperLoRA, parameter-efficient portrait adaptation.ComfyUI-PhotoMaker-Plus(commit459f5c00) — PhotoMaker Plus personalized portrait generation.ComfyUI_StoryDiffusion(commitc228e192) — StoryDiffusion coherent story image generation.comfyui-SelfLift(commit19ec5405) — SelfLift nodes.
- Tagging / Captioning
ComfyUI-Miaoshouai-Tagger(commitbb92078e) — Image tagging.ComfyUI-JoyCaption(commita0e9f0a1) — LLaVA-based image captioning, supports batching and GGUF.
- Background Removal / Upscale / Utility
ComfyUI_BiRefNet_ll(commit5443a2aa) — BiRefNet background removal.ComfyUI-Flowty-LDSR(commitf9501ecc) — LDSR super-resolution.ComfyUI-enricos-nodes(commite8752264) — Compositor node experiments.ComfyUI-Kolors-MZ(commit43ec2701) — Kolors native sampler implementation.ComfyUI-ResAdapter(commit0567e8a0) — ResAdapter domain-prompt resolution adapter LoRA.
Breaking Changes
Updated custom nodes:ComfyUI-KJNodesc9869ea→d3cfe21ComfyUI-MiniMax-H3-Turbo55fee864→4274783comfyui-minimax-h3-audio-T8ad43472→4a9269fComfyUI_MiniMaxH3_Director8418d4f→04abbafcomfyui-minimax-h3-blockcache-T828eda986→36336dcComfyUI-MAINodes509aefc→f4868b4ComfyUI-YCNodes-MiniMax-H314620729→efbe83fComfyUI-H3-Motion-Contexta12a50c→5335715ComfyUI-SeedVR2_VideoUpscaler→f4de6e2ComfyUI-qwenmultianglec264343→bf8d032
ComfyUI-LatentSyncWrapperComfyUI-QualityOfLifeSuit_Omar92
Usage Notes
- Restart the environment / refresh the page to apply
- Workflows referencing
ComfyUI-LatentSyncWrappernodes must migrate to other lip-sync solutions
September 11, 2026
Summary: This update adds 29 custom nodes and updates 7. Highlights: 3D generation (Hunyuan3D / Zero123), video matting & noise control, MiniMax-H3 extension suite, Flux stylization / face swap, with ComfyUI core upgraded to v0.34.0.Core Update
- ComfyUI version: v0.33.0 (e3a6631) → v0.34.0 (d07ab68)
- Effective: immediately (refresh the page)
New Capabilities
New custom nodes added (29 total):- 3D Generation
ComfyUI-Hunyuan3d-2-1(commite0ddbbb) — Hunyuan3D v2.1 image-to-3D assets.ComfyUI-Hunyuan3DWrapper(commit2609efa) — Hunyuan3D 3D generation wrapper.StableZero123-comfyui(commit49c4837) — Zero123 3D novel view synthesis.
- Depth / Pose
ComfyUI-DepthCrafter-Nodes(commitf469414) — DepthCrafter consistent long-video depth estimation.ComfyUI-Depth-Pro(commit2164e20) — Apple ml-depth-pro monocular depth.ComfyUI-Marigold(commit2adfec7) — Marigold diffusion-based depth estimation.ComfyUI_3dPoseEditor(commit582c5b1) — 3D pose editor.
- Video Processing
ComfyUI-WanMoeKSampler(commitee8fa8f) — WanVideo Moe KSampler modification.ComfyUI-Video-Matting(commitdd5ff37) — Robust Video Matting (RVM).ComfyUI-VideoNoiseWarp(commitb08b263) — Video noise warp nodes.ComfyUI-PainterLTXV2(commitb316d73) — LTXV audio-video split sampling & latent prep.HandFixer(commit4817d68) — AI hand repair workflow nodes.
- MiniMax-H3 Extensions
ComfyUI-MiniMax-H3-Hybrid(commitb368af7) — MiniMax H3 Hybrid nodes.ComfyUI-H3-Motion-Context(commita12a50c) — H3 Motion Context.ComfyUI-H3-Context-Noise(commit7e55312) — H3 Context Noise.
- Flux / Image Editing / Background Removal
x-flux-comfyui(commit0032855) — XLabs-AI Flux style LoRA + ControlNet.ComfyUI-BRIA_AI-RMBG(commit827fcd6) — BRIA RMBG background removal.ComfyUI-ACE_Plus(commit616c478) — ACE Plus workflow nodes.Lucy-Edit-ComfyUI(commit2bec612) — Lucy Edit image editing.comfyui-krea2edit(commit86f886d) — Instruction-based, identity-preserving image editing for Krea 2.
- Portrait / Face Swap
comfyui-instantId-faceswap(commitd125a2a) — Face swap based on InstantID.CharacterFaceSwap(commitf395496) — Character face swap nodes.
- Utility / Other
ComfyUI_CineStyle(commitf3c3f0d) — Cinematic color grading LUT film emulation.ComfyUI-Panorama-Stickers(commitaf0357f) — Panorama sticker nodes.ComfyUI_preview360panorama(commit2a44014) — Interactive 360 panorama preview.ComfyUI-ZML-Image(commit9760a4e) — Text/image utility nodes.ComfyUI-QualityOfLifeSuit_Omar92(commit793b6d2) — Quality of life helper nodes.SDVN_Comfy_node(commit707f799) — SDVN node pack.ComfyUI_SLK_joy_caption_two(commita61f9a9) — JoyCaption2 image captioning.
Bug Fixes
- Fixed MinimaxH3LatentUpscaler3D node rename (MinimaxH3LatentUpscalerNode3D → MinimaxH3LatentUpscaler3D) and V3 IO schema adaptation
Breaking Changes
Updated custom nodes (including core):ComfyUI e3a6631(v0.33.0) → d07ab68 (v0.34.0)ComfyUI-KJNodes3f20054 → c9869eaComfyUI-DepthAnythingV2d505cbc → 5531878WhatDreamsCost-ComfyUIda7c5d3 → a3c809cComfyUI-Anima-LLLitec8ad372 → b7495bdcomfyui-minimax-h3-audio-T8dc50c2e → ad43472Comfyui_Minimax_h3_latent_Upscaler59f399f → d7c01b9
Augest 26,2026
Summary: This update adds 22 custom nodes, updates 2. Headline: MiniMax-H3 video suite (9 nodes) plus Wan/Hunyuan/CogVideoX video tooling, with ComfyUI upgraded to v0.33.0.Core Update
- ComfyUI version:
517799b→e3a6631(v0.33.0) - Effective: immediately (refresh the page)
New Capabilities
New custom nodes added: MiniMax-H3 video suiteComfyUI-MiniMax-H3-Turbo(commit55fee86) — Turbo-accelerated MiniMax-H3 text-to-video.ComfyUI-YCNodes-MiniMax-H3(commit1462072) — YCNodes MiniMax-H3 video generation nodes.ComfyUI_MiniMaxH3_Director(commit8418d4f) — MiniMax-H3 Director-mode video generation.comfyui-minimax-h3-audio-T8(commitdc50c2e) — MiniMax-H3 audio gen + Skin Finish post-processing.comfyui-minimax-h3-blockcache-T8(commit28eda98) — MiniMax-H3 with block cache for faster inference.Comfyui_Minimax_h3_latent_Upscaler(commit59f399f) — Neural latent upscaler for H3, skips VAE decode/encode.ComfyUI-SolAttn_triton(commit842c4ea) — SolAttn triton kernels for faster attention.ComfyUI-MAINodes(commit509aefc) — Contact-Sheet diffusion + Motion Lab (test-time de-roping).ComfyUI-CacheDiT(commit1d92bbd) — Cache-DiT acceleration node.
ComfyUI-WanVideoStartEndFrames(commit340b4f0) — WanVideo first/last frame control.ComfyUI-wanBlockswap(commit5fa2ec0) — WanVideo blockswap for lower VRAM usage.ComfyUI-WanVideoKsampler(commite92a703) — WanVideoKsampler node (companion to wanBlockswap).ComfyUI-HunyuanVideoMultiLora(commit9e18b97) — Multi-LoRA loading for HunyuanVideo.ComfyUI-CogVideoXWrapper(commitfdb8abd) — CogVideoX video generation wrapper.ComfyUI-CogVideoX-MZ(commitecdaefe) — CogVideoX-MZ video generation.ComfyUI-PainterI2VforKJ(commit56f6e8a) — PainterI2V image-to-video.ComfyUI-MochiEdit(commit5929b50) — Edit videos with the Mochi model.
comfyui-liveportrait(commitffde131) — LivePortrait portrait animation (KwaiVGI).ComfyUI-HY-Motion1(commit1342cd8) — HY-Motion1 video motion control.ComfyUI-mesh2motion(commit11fe6b7) — mesh2motion 3D motion transfer.ComfyUI-SCAIL2-Easy(commitf4ab426) — SCAIL-2 simplified nodes.comfyui-dynamicprompts(commit3f2fff3) — Dynamic prompt generation (random/combinatorial/Jinja).
Bug Fixes
- Fixed onnxruntime
- Fixed ComfyUI-LTXVideo runtime issue (commit
531512f→15d09ab) - Fixed comfyui-minimax-h3-blockcache-T8 runtime issue
- Fixed LoraLoaderBypassModelOnly configuration
- Repaired comfy-kitchen (
0.2.26→0.2.31) - Upgraded SageAttention to
2.2.0for compatibility
Breaking Changes
Updated custom nodes:ComfyUI-KJNodesa8fd39c→3f20054ComfyUI-LTXVideo531512f→15d09ab