My ComfyUI workflows for using Wan 2.2
This workflows are used by me to create my art.
They are optimized for my checkpoints and created of my latest knowledge to enhance the outcome.
"If this workflow leveled up your day, I'd purr-eciate a like! 😻"
Versions & Information👇👇👇👇👇👇👇👇
👉 Please read below and the file descriptions "About this version" for more info's.
🎬🐈Click me! HowTo + Video
💡Some WAN 2.2 versions use high+low checkpoints, other like S2V use only a single checkpoint - make sure to read the descriptions and use the correct checkpoints.
⚠️ Do not use the workflows with the "Nodes 2.0 beta" from ComfyUi or it will mess up things.
👇👇👇👇👇👇👇👇
What you get from the comfy workflows:
♨️ Easy controls
✅ As less as possible dependencies
🪧 Detailed documentation
⛓️ Highly automatic logic
✨ Optimized results
🎬 Fully automated resolution logic
🔖 Bookmark-Shortcuts with number keys
Types of workflows
FastFidelity C-AiO
🖼️ I2V and FLF2V
🧩 Automatic aspect‑ratio calculation and fitting
✨ Multiple Upscalers
Torchlanc (very fast, color correct, low VRAM)
Upscale with Model (additional detail, high quality)
RTX Super Resolution (ultra fast, very accurate)
🤝 Video resolution matching - Fully automatic scaling and resolution calculations
🕠 Length automation - Fully automatic calculation of frame count
🫥 Watermark option
🧮 Color match feature
👾 MiniMeme feature - Create small gif's
🎭 NAG - Negative prompting with CFG1
🪄 Interpolation feature
🌀 Perfect loop feature
🃏 Last Frame Extraction
🔖 Bookmark-Shortcuts - with number keys
FastFidelity C-SVI
🖼️ SVI (SVI 2.0 up to 10 samplers)
🧩 Automatic aspect‑ratio calculation and fitting
✨ Ultra‑fast, color‑correct upscaling (torchlanc) OR Upscale with Model
🤝 Video resolution matching - Fully automatic scaling and resolution calculations
🕠 Length automation - Fully automatic calculation of frame count
🫥 Watermark support
🪄 Interpolation feature
🔖 Bookmark-Shortcuts - with number keys
Swarm Basic
Absolute basic SwarmUI preset
Not recommended over ComfyUI for video generation
Backend Test
If this does not create the example video your ComfyUI backend is broken.
If this works, but your other workflow not, the other workflow is broken or missing dependencies
🩻 Known issues and advice's
⚠️ Some workflows may set on webp av1 encoding (VHS node) - If your computer/setup missing drivers use any other like H265 or H264!
Install ffmpeg!
Update Comfyui and custom_nodes!
Update pytorch 2.9+cu128 or higher
Make sure to read where files/models should be placed inside the workflow
Check if the filepath for model/clip/vae match your system like Linux/Windows
The plugin ComfyUI-DD-Translation can break node connection (avoid)
All older Versions are available inside my GitHub Repo.
Spacial thanks to @Abyss_Games for a really good idea and help with the "loop"!
I got some really good ideas from @Gladas workflows!
YOU are responsible for outputs as always! If you make ToS violating content and I get aware I WILL report this.
Description
Minor optimisations and simplifications
Adjusted to use the baked in VAE and CLIP from my checkpoint "RadiantCrush"
1-shot generating, no extra VAE and CLIP needed
Requirements:
FAQ
Comments (28)
I tried it with GGUF + (4step 1cfg loras), but even though the outputs were fast, I couldn't find the exact consistent settings. So I was thinking of trying it with the safetensors model you use, but I couldn't find it. What model are you using? “wan2_2-I2V-L-LS-DaSiWa-nM_RadiantCrush.safetensors” is the name of the model selected in the workflow, I think, but I couldn't find it online. Maybe changed the name, I'm not sure.
It is this model: https://civitai.com/models/1981116
Okay, this question might be a bit stupid, but should I load the high-noise version and low-noise version separately on the checkpoint loading nodes and enable both nodes at the same time, right?
You have to use both, all enabled. As set by default.
@darksidewalker Okay, still a pretty dumb question: If I need to use LoRA, do I need to put the high-noise LoRA into the first LoRA loader, and the low-noise one into the second LoRA loader? Additionally, connect the CLIP output port of the high-noise checkpoint to the CLIP input port of the first LoRA loader? If my description is wrong, what should be done correctly?
@NAGISAZUKI No new connections needed, just add the loras to the named lora stackers.
@darksidewalker but if I didn't make a new connections,I will get this erro:Prompt execution failed
Prompt outputs failed validation: Power Lora Loader (rgthree): - Required input is missing: clip Power Lora Loader (rgthree): - Required input is missing: clip
i import the workflows and some node are missing, i dont know where to install/find the missing node to install it manually
I forgot to add this in the version of the last WF, added it . Also mentioned it on the main page.
@darksidewalker thank you, sorry for the trouble still newbie in comfyui
Im not sure why but it keep getting stuck on Ksampler 2 or Vae decoder, I have 3090 so I assumed the specs would be perfectly fine. The first Ksampler only takes 200s to run.
Same here. Got stuck in the second Ksampler.
Any error or just oom?
You could try to disable all VRAM cleanup nodes.
@darksidewalker No errors, just taking incredibly long to finish/just is stuck. Maybe oom, at 656x960 with 32g Vram I assumed it would be okay. Perhaps I will try with lower resolution. it also doesnt seem to reach any vram clean up nodes before its stuck
@Sachiw I release a new version, you could have a try with the standard resolution.
@darksidewalker Thank you, will try that in a bit
@Sachiw I would guess you have to less normal system RAM and getting oom or extreme long times because of virtual memory swap.
@darksidewalker i do have 64g regular ram as well, though it might not be all free. Ill try to run it after restarting to see if that fixes it.
@darksidewalker Okay yeah, after clearing out systemRAM+restart it ran successfully. Thank you so much for the help, and the output was great! WF was easy to setup as well!
hello,
thank you for this workflow on the other hand the Frames, width, height, steps fields are empty. we cannot indicate anything as a value inside.
how to correct the problem?
This are standard comfy nodes, this may be a problem with your comfy. Maybe update or fresh install.
I installed swarmUI and I have the same problem however I can modify it by transferring the workflow I will try from here!
I also use Swarmui and just use the workflow there, because it can do some things swarm is not, but swarm also just works for me, without the upscale part.
@darksidewalker
I am sorry to bother you again but with SwarmUI I have this error message precisely with the UPSCALE_method whether it is disabled or activated with nearest-exact or other:
ComfyUI execution error: ImageScale.upscale() missing 1 required positional argument: 'upscale_method'
I deactivated the upscaler and was able to launch a generation.
I have a 4070 super 12Go for now it started but it’s stopped at
ComfyUI execution error: KSamplerAdvanced.sample() missing 1 required positional argument: 'add_noise'
@aleode764 you somehow corrupted your comfy/swarm installation or the workflow. The errors are not a problem of the uploaded workflow itself. You should redownload the workflow or/and fresh install comfy/swarm.
@darksidewalker
I gave up, I downloaded a portable version of comfyUI and I have the same problems.
I will see with another workflow. my swarmUI installation was done today. it’s okay I’ll test your models later.
thank you very much for answering me anyway.
I reworked a workflow taking into account your parameters and it works very well.
Bravo thank you very much for your models.
https://civitai.com/posts/23693713
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.