Skip to content
WorkflowListedExternal
Open source

Wan2.2 I2V/14B/GGUF KIJAI WF with For loop lightxv2. — v2.0

This is a workflow base from kijai wanvideowrapper. very user friendly. I added for loop and also added lora for each video step. I Took idea from some other workflow then did it with kijai node and added some note...

Sourcecivitai.com/models/1845620
TaskMultiCapabilitiesImage to image · UpscalingFamilyWanEstimated VRAMNot yet characterizedBgrade

Runtime profile

Access
External workflow
Status
Listed
Source
Civitai
Updated
Oct 2, 2026

Source description

This is a workflow base from kijai wanvideowrapper. very user friendly. I added for loop and also added lora for each video step. I Took idea from some other workflow then did it with kijai node and added some note in the workflow. I cleaned it up also. Im using an rtx 3080 so everything is set for 10 to 12GB. but you can tweak thing as you want.

For those who doesn't know what for loop it is basically longer video and it take the last frame of your first generation and continue your video from this last frame etc... so you can get long video.

EDIT: I saw that some people was still downloading my old version and had some question so i updated it and made it a bit more clear.

What you need

Model files
16.9 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

Requirements

Models

9

Node packs

9

  • comfyui_layerstyle

    Found

    Node pack · Registry

  • comfyui_memory_cleanup

    Found

    Node pack · Registry

  • comfyui-easy-use

    Found

    Node pack · Registry

  • comfyui-frame-interpolation

    Found

    Node pack · Registry

  • comfyui-gguf

    Possible

    Node pack · Registry

  • comfyui-kjnodes

    Found

    Node pack · Registry

  • comfyui-videohelpersuite

    Found

    Node pack · Registry

  • ComfyUI-WanVideoWrapper

    Found

    Node pack · Registry

  • rgthree-comfy

    Found

    Node pack · Registry

AI-generated commentary

AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.

Purpose

  • Use this Wan Video 2.2 I2V-A14B workflow for image-to-video generation; it can extend a clip into a longer video by continuing from each generation’s last frame.

Flow

  • The workflow adds a LoRA at each video step.

Inputs and outputs

  • For each continuation, the previous generation’s last frame is used as the next video step’s input.

Setup

  • View or download the workflow on Civitai: https://civitai.com/models/1845620.

Dependencies

  • Node packs named for the workflow: comfyui-easy-use, rgthree-comfy, ComfyUI-WanVideoWrapper, comfyui-gguf, comfyui-frame-interpolation, comfyui-kjnodes, comfyui_layerstyle, comfyui_memory_cleanup, and comfyui-videohelpersuite.

  • Named LoRA files: Wan_2_2_I2V_A14B_HIGH_lightx2v_4step_lora_v1030_rank_64_bf16.safetensors, 23High noise-Cumshot Aesthetics.safetensors, 56Low noise-Cumshot Aesthetics.safetensors, and lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank256_bf16.safetensors.

  • Other named files: RealESRGAN_x2plus.pth, umt5xxl-encoder-q6_k.gguf, wan_2.1_vae.safetensors, wan2.2_i2v_high_noise_14B_Q4_K_S.gguf, and wan2.2_i2v_low_noise_14B_Q4_K_S.gguf.

Configuration

  • The author says they use an RTX 3080 and set the workflow for 10–12 GB; settings can be adjusted.

Cautions

  • Before setup, check which package is meant by comfyui-gguf, and confirm the intended files named 23High noise-Cumshot Aesthetics.safetensors, 56Low noise-Cumshot Aesthetics.safetensors, and umt5xxl-encoder-q6_k.gguf.

    Suggestion · not verified

Does this need to be edited?

Sign in to send an edit request.

Sources

1 source
CivitaiPrimary
civitai.com/models/1845620

Verified Jul 19

Source excerpts

1 excerpt
NoteJul 19civitai.com/models/1845620

Source context: 1890 downloads · Type Workflows · Base model Wan Video 2.2 I2V-A14B