Simple workflow. Wan 2.1 for the 3060 RTX 12 GB VRAM on 32 GB System RAM Windows 10 PC.
created 15 days ago
video
12 nodes

3.2k

710

Credits
not sure who it came from originally but it just worked off the shelf once I downloaded all the models
Outputs
example image output 0 for workflow.
example image output optimized_comfy_workflows_user_uploads/8de910f5-6e22-4819-b7a0-f5b5dd5f4ce5/assets/example woman.webp for workflow.
Description

Posting this i2V WAN 2.1 workflow, because it works on mine, and a lot of people are getting too complex with the workflows and it aint working. KISS to get it working. Dont expect super high quality in 30 seconds of rendering time, but its about getting it working first then building up tricks from there.

the attached file of the woman and the created the mp4 are what I used and made with the workflow and took I think 16 minutes with those settings. Try smaller like 416 x 416 or whatever to test it at first. The result shared here in the mp4 is a big jiggy but that is coz its 16fps I then use other software to interpolate it to 24 fps or more which is quicker than trying to do it on the 3060 in comfyui

presuming you have 12GB VRAM I am on windows 10. I have run every WAN model without issue so far other even 15 GB in size, and other than the time it takes and lower quality in some cases, not had any OOMs. I definitely recommend the city69 GGUF quant models and go for sizes that will fit your VRAM. also use super simple workflows. all the complex ones are just slowing my machine down tbh. its too early for it to be tweaked for extra stuff imo e.g. teacache just slowed my WAN 2.1 i2v rendering down. (early March 2025). Though that will change as the coders work on improving it all so this post might already be obsolete.

if you are getting OOM errors...

  1. first I would make sure nothing else is using the VRAM that shouldnt be before you get started.

  2. then make sure Comfyui and NVDIA drivers and python and everything is up to date where you want it to be.

  3. then when booting up comfyui check the command window and note down all the errors in a notepad file and go through each looking at how to fix them (i usually copy and search a generic part of the error and look for github issues where someone had the same problem). if its not fixable move to the next.

  4. then if its still crashing after all that, detail exactly what info you are getting around that, when it does, and seek out help on google related to what you find to fix each problem.

Built-in nodes
WanImageToVideo
Custom nodes
CLIPTextEncode
CLIPVisionEncode
KSampler
CLIPLoader
VAELoader
CLIPVisionLoader
VHS_VideoCombine
VAEDecode
LoadImage
UnetLoaderGGUF
1
0
0
0