Dear SeaArt ComfyUI users,
Thank you for your continued attention and support! In this update, we will bring feature optimizations and bug fixes based on valuable community feedback.
Feature Updates
1. Task runtime is now displayed during execution, allowing users to track progress in real time.
2. When downloading video generation results, the video file will be automatically named after the current workflow for easier management and retrieval.
3. The presentation of tutorial content has been optimized for a better learning experience.
4. Added Creation Assistant, which can explain node functions and provide basic answers about workflows and ComfyUI knowledge.
5. Support for hunyuan image 2.1 generation has been added in this upgrade.
Workflow link: [View here]
Article link: [View here]
Node Updates
New Nodes
ComfyUI has undergone a major upgrade and now supports a large number of new nodes:
LatentConcat, LatentCut, SamplerER_SDE, SamplerSASolver, SamplingPercentToSigma, ImageAddNoise, ImageStitch, ResizeAndPadImage, ImageFlip, ImageScaleToMaxDimension, TrainLoraNode, SaveLoRANode, LoadImageSetFromFolderNode, LoadImageTextSetFromFolderNode, LossGraphNode, ModelMergeCosmosPredict2_2B, ModelMergeCosmosPredict2_14B, ModelMergeQwenImage, SaveAudioMP3, SaveAudioOpus, RecordAudio, EmptyHunyuanImageLatent, HunyuanRefinerLatent, FluxKontextMultiReferenceLatentMethod, SkipLayerGuidanceDiTSimple, CosmosPredict2ImageToVideoLatent, Wan22FunControlToVideo, WanSoundImageToVideo, WanSoundImageToVideoExtend, CFGNorm, APG, WanCameraEmbedding, TCFG, ContextWindowsManual, WanContextWindowsManual, EmptyChromaRadianceLatentImage, ChromaRadianceOptions, ModelPatchLoader, QwenImageDiffsynthControlnet, USOStyleReference, EasyCache, LazyCache, AudioEncoderLoader, AudioEncoderEncode, OpenAIChatNode, OpenAIInputFiles, OpenAIChatConfig, MinimaxHailuoVideoNode, Veo3VideoGenerationNode, FluxKontextProImageNode, FluxKontextMaxImageNode, ByteDanceImageNode, ByteDanceImageEditNode, ByteDanceSeedreamNode, ByteDanceTextToVideoNode, ByteDanceImageToVideoNode, ByteDanceFirstLastFrameNode, ByteDanceImageReferenceNode, StabilityTextToAudio, StabilityAudioToAudio, StabilityAudioInpaint, RunwayFirstLastFrameNode, RunwayImageToVideoNodeGen3a, RunwayImageToVideoNodeGen4, RunwayTextToImageNode, TripoTextToModelNode, TripoImageToModelNode, TripoMultiviewToModelNode, TripoTextureNode, TripoRefineNode, TripoRigNode, TripoRetargetNode, TripoConversionNode, MoonvalleyImg2VideoNode, MoonvalleyTxt2VideoNode, MoonvalleyVideo2VideoNode, Rodin3D_Regular, Rodin3D_Detail, Rodin3D_Smooth, Rodin3D_Sketch, GeminiNode, GeminiImageNode, GeminiInputFiles, ViduTextToVideoNode, ViduImageToVideoNode, ViduReferenceVideoNode, ViduStartEndToVideoNode, GetLatentSizeAndCount, LoadVideosFromFolder, DiffusionModelSelector, LazySwitchKJ, TorchCompileModelQwenImage, LoraReduceRankKJ, WanVideoEncodeLatentBatch, WanVideoTextEncodeCached, WanVideoAddExtraLatent, WanVideoScheduler, WanVideoAddStandInLatent, WanVideoAddControlEmbeds, WanVideoAddMTVMotion, WanVideoRoPEFunction, WanVideoAddPusaNoise, FantasyPortraitModelLoader, FantasyPortraitFaceDetector, WanVideoAddFantasyPortrait, LandmarksToImage, Wav2VecModelLoader, MultiTalkSilentEmbeds, WanVideoLoraSelectByName, WanVideoExtraModelSelect, WanVideoLatentReScale, CreateScheduleFloatList, WanVideoSigmaToStep, NormalizeAudioLoudness, WanVideoPassImagesFromSamples, QwenLoader, WanVideoPromptExtender, WanVideoPromptExtenderSelect, DownloadAndLoadNLFModel, NLFPredict, DrawNLFPoses, LoadVQVAE, MTVCrafterEncodePoses, WanVideoAddS2VEmbeds
Custom Node Upgrades
The following custom nodes have been upgraded to the latest version:
- ComfyUI-WanVideoWrapper
- ComfyUI-KJNodes
- ComfyUI-Advanced-ControlNet
- ComfyUI-TeaCache
Bug Fixes
The following issues have been identified and are currently being fixed. The corresponding fixes will be released soon:
1. The issue of embeddings not working properly. Fixed
2. Some model display issues (such as with Nunchaku and WanVideoWrapper nodes). Fixed
3. The problem where the Load Image node cannot open the right-click menu after loading an image. Fixed
NEW Bug Fix Log
- Issue: The number of LORA usages was unintentionally restricted for ComfyUI users.
- Cause: A system bug mistakenly triggered the LORA usage limit.
- Progress: The bug has been identified and the development team is working on a fix.
- Affected Users: All ComfyUI users.
- Note: We will announce as soon as the fix is released. Thank you for your understanding and support.
---
SeaArt will continue to optimize the ComfyUI experience. We welcome your feedback and suggestions.
Thank you for your attention and support!
