<h2 id="svi-extend">SVI Extend</h2><a target="_blank" rel="ugc" href="https://github.com/vita-epfl/Stable-Video-Infinity/tree/svi_wan22">https://github.com/vita-epfl/Stable-Video-Infinity/tree/svi_wan22</a>Create videos and extend them seemlessly using SVI. Following SVI LoRAs are mandatory:HIGH: <a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/blob/main/LoRAs/Stable-Video-Infinity/v2.0/SVI_v2_PRO_Wan2.2-I2V-A14B_HIGH_lora_rank_128_fp16.safetensors">https://huggingface.co/Kijai/WanVideo_comfy/blob/main/LoRAs/Stable-Video-Infinity/v2.0/SVI_v2_PRO_Wan2.2-I2V-A14B_HIGH_lora_rank_128_fp16.safetensors</a>LOW: <a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/blob/main/LoRAs/Stable-Video-Infinity/v2.0/SVI_v2_PRO_Wan2.2-I2V-A14B_LOW_lora_rank_128_fp16.safetensors">https://huggingface.co/Kijai/WanVideo_comfy/blob/main/LoRAs/Stable-Video-Infinity/v2.0/SVI_v2_PRO_Wan2.2-I2V-A14B_LOW_lora_rank_128_fp16.safetensors</a><ul><li>switch between default behaviour, anchor_samples and end_frames within the same subgraphs</li><li>connect an image to a part and enable the respective toggles to use end_frames or anchor_samples</li></ul>NEW! v3Extend existing videos using <a target="_blank" rel="ugc" href="https://github.com/wallen0322/ComfyUI-Wan22FMLF">https://github.com/wallen0322/ComfyUI-Wan22FMLF</a><ul><li>enable "video extension" toggle inside the settings</li><li>uses source video resolution by default<ul><li>rescale video using the megapixel slider by enabling "video rescale" toggle</li></ul></li><li>use included version of the nodes from inside .zip or download the latest version straight from the git if issues arise</li></ul>More info inside the workflow.<h2 id="aio-i2v+t2v">AIO i2v+t2v</h2>All in One workflow for for basic WAN 2.2 video generation.Following features included:<ul><li>Switch seemlessly between 2 and 3 sampler solutions</li><li>Toggle between i2v or t2v</li><li>Postprod<ul><li>Facedetailer<ul><li>uses t2v Model + LoRA for inpainting - resources needed included in workflow</li></ul></li><li>Toggle between GIMM VFI and RIFE VFI Interpolation</li><li>Upscale<ul><li>Tensorrt Upscale with Model</li><li>Basic Video Upscale with Model</li><li>RTX Video Super Resolution Upscale (insanely fast for decent quality)</li></ul></li><li>Frame Clipper</li><li>Seamless Loops using custom RIFE nodes <a target="_blank" rel="ugc" href="https://github.com/Artificial-Sweetener/comfyui-WhiteRabbit">https://github.com/Artificial-Sweetener/comfyui-WhiteRabbit</a></li></ul></li></ul><h2 id="upscale-+-interpolate">Upscale + Interpolate</h2>I recommend using this workflow instead of upscaling with the generating workflows, since you never really know what kind of results you get, ending up upscaling a bad video and wasting time. I included toggles so you can't use multiple interpolation or upscale nodes at once by mistake.This includes:<ul><li>WAN Facedetailer<ul><li>use any WAN 2.2 T2V low model + the following T2V LoRA:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Lightx2v/lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank128_bf16.safetensors">https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Lightx2v/lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank128_bf16.safetensors</a></li></ul></li><li>lower resolution from 768 to 512 if you have VRAM issues</li><li>Put the following file into "ComfyUI\models\ultralytics\bbox":<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Anzhc/Anzhcs_YOLOs/blob/main/Anzhc%20Face%20seg%20640%20v3%20y11n.pt">https://huggingface.co/Anzhc/Anzhcs_YOLOs/blob/main/Anzhc%20Face%20seg%20640%20v3%20y11n.pt</a></li></ul></li></ul></li><li>WAN Refiner (massive VRAM cost)<ul><li>increase denoise if you want more inpainting</li></ul></li><li>Sharpen, Gamma, Brightness and Contrast control</li><li>Frame clipper (remove unwanted frames at the start and/or end)</li><li>GIMM VFI + RIFE VFI interpolation (I recommend GIMM VFI, much higher quality but also much slower)<ul><li>see the following for a massive RIFE speedup<ul><li><a target="_blank" rel="ugc" href="https://github.com/Fannovel16/ComfyUI-Frame-Interpolation/pull/102">https://github.com/Fannovel16/ComfyUI-Frame-Interpolation/pull/102</a></li><li><a target="_blank" rel="ugc" href="https://github.com/Fannovel16/ComfyUI-Frame-Interpolation/blob/fa2f594ff0ecd4373b07aad87f1f685a0b8a771e/vfi_models/rife/__init__.py">https://github.com/Fannovel16/ComfyUI-Frame-Interpolation/blob/fa2f594ff0ecd4373b07aad87f1f685a0b8a771e/vfi_models/rife/__init__.py</a></li></ul></li></ul></li><li>Tensorrt Upscale + Basic Video Upscale<ul><li>both use basic image upscaling models</li><li>Tensorrt (faster than Basic Video Upscale) with AnimeSharp4x is recommended for anime</li></ul></li><li>RTX Video Super Resolution Upscale<ul><li>insanely fast</li><li>decent quality</li></ul></li><li>FlashSVR + SeedVR2<ul><li>experimental</li><li>Video upscale models that are more intricate than basic image upscaling models</li><li>haven't had great results for anime yet</li><li>takes a LOT longer</li></ul></li><li>Saving last frame for manual extensions</li></ul><h3 id="mmaudio">mmaudio</h3><ul><li>added Audio combine node<ul><li>combine audio from an existing video with the generated audio on top</li><li>generate nsfw audio with the nsfw model and then combine that video with another generated audio track from the base model for background noises</li></ul></li><li>removed interpolation for easier and faster audio generation - you have the following options:<ul><li>upload raw unupscaled video to MMAudio Video node and upscaled video to Combine video node</li><li>upload upscaled video to both nodes but lower custom_width and custom_height of the MMAudio video node to about half for faster generation and to prevent VRAM issues</li><li>upload raw video to both nodes and upscale afterwards</li></ul></li></ul>Inspired by <a target="_blank" rel="ugc" href="https://civarchive.com/models/2137833">https://civarchive.com/models/2137833</a>Following resources necessary (ComfyUI\models\mmaudio):<a target="_blank" rel="ugc" href="https://huggingface.co/phazei/NSFW_MMaudio/resolve/main/mmaudio_large_44k_nsfw_gold_8.5k_final_fp16.safetensors?download=true">https://huggingface.co/phazei/NSFW_MMaudio/resolve/main/mmaudio_large_44k_nsfw_gold_8.5k_final_fp16.safetensors?download=true</a><a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MMAudio_safetensors/resolve/5984623e6b436818c6ff287ef6eec93e3e05aa3f/mmaudio_vae_44k_fp16.safetensors">https://huggingface.co/Kijai/MMAudio_safetensors/resolve/5984623e6b436818c6ff287ef6eec93e3e05aa3f/mmaudio_vae_44k_fp16.safetensors</a><a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MMAudio_safetensors/resolve/main/mmaudio_synchformer_fp16.safetensors">https://huggingface.co/Kijai/MMAudio_safetensors/resolve/main/mmaudio_synchformer_fp16.safetensors</a><a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MMAudio_safetensors/resolve/5984623e6b436818c6ff287ef6eec93e3e05aa3f/mmaudio_vae_44k_fp16.safetensors">https://huggingface.co/Kijai/MMAudio_safetensors/resolve/5984623e6b436818c6ff287ef6eec93e3e05aa3f/mmaudio_vae_44k_fp16.safetensors</a>

WAN 2.2 i2v ComfyUI Workflow + SVI Extend (FLF2V) + Upscale

Added GIMM VFI Interpolation for postproduction<ul><li>toggle between RIFE and GIMM VFI</li></ul>Added Motion Scale <a target="_blank" rel="ugc" href="https://github.com/shootthesound/comfyUI-LongLook">https://github.com/shootthesound/comfyUI-LongLook</a><ul><li>speed up or slow down motion</li><li>values between 1.2-1.5 recommended</li></ul>Added rudimentary T2V Support<ul><li>toggle T2V Latent on the Backend subgraph and adjust T2V width/height</li></ul>Added Frame Clipper<ul><li>clip start/end frames</li><li>Useful in combination with <a target="_blank" rel="ugc" href="https://civitai.com/models/2264812?modelVersionId=2549297">https://civitai.com/models/2264812?modelVersionId=2549297</a></li></ul>Added WanCutLastSlot<ul><li>remove unwated white flashing frames when using FLF2V</li></ul>

Files with SHA256: 59ef744ed37ee43fd85fec945671806ac6eb01e65175187fd45550f260eb0fb5