Wan 2.2 Video Upscale — video upscale v1.1

The “Wan 2.2 Video Upscale — video upscale v1.1” workflow is for upscaling video with a Wan 2.2 Low-Noise model; it lists LoadImage and VHS_LoadVideo as image and video inputs and VHS_VideoCombine for video output.

The workflow has 43 nodes in one file, “Wan Upscale v1.1.json.”

Listed node types include WanImageToVideo, VAEEncode, VAEDecodeTiled, KSamplerAdvanced, frame-selection nodes, and VHS_VideoCombine.

LoadImage and VHS_LoadVideo are the listed image and video input nodes.

For output handling, the workflow lists VHS_VideoCombine, GetImageRangeFromBatch, and FinalFrameSelector.

The selected release is “video upscale v1.1,” and the named base model is “Wan Video 2.2 I2V-A14B.”

The workflow asset is an archive named wan22VideoUpscale_videoUpscaleV11.zip.

The listed model files are a checkpoint, LoRA, text encoder, and VAE: DasiwaWAN22I2V14BLightspeed_synthseductionLowV9.safetensors; LOW_SVI_v2_PRO_Wan2.2-I2V-A14B_LOW_lora_rank_128_fp16.safetensors; umt5_xxl_fp8_e4m3fn_scaled.safetensors; and wan_2.1_vae.safetensors.

Listed custom nodes are comfyui_essentials, comfyui-easy-use, comfyui-kjnodes, comfyui-logicutils, ComfyUI-MediaMixer, comfyui-videohelpersuite, rgthree-comfy, and scale-image-to-total-pixels-advanced.

For image and frame handling, the listed node types include ImageScaleToTotalPixelsX, RepeatImageBatch, BatchImagesNode, GetImageRangeFromBatch, and FinalFrameSelector.

The listed node types also include CLIPTextEncode, LoraLoaderModelOnly, ModelSamplingSD3, and ColorMatch.

LOW_SVI_v2_PRO_Wan2.2-I2V-A14B_LOW_lora_rank_128_fp16.safetensors is listed as manual and has no source URL; verify how you will obtain it before use.

Vorschlag · nicht geprüft

No license is listed for this workflow, and no full model artifact is included; check usage terms and locate the four model files before use.

Vorschlag · nicht geprüft