WorkflowListedExternal

Wan2.2 GGUF Workflow test — Wan2.2 GGUF WF

Wan2.2 GGUF WF

Sourcecivitai.com/models/1818827

TaskText to videoFamilyWanVersionRelease 2058273 — want2v_2.2_00001.jsonEstimated VRAMEstimate unavailableBgrade

Runtime profile

Access: External workflow
Status: Listed
Version: want2v_2.2_00001.json
Source: Civitai
Updated: Jul 19, 2026

Source description

WAN2.2 GGUF Workflow 1st impressions

My PC : 64 GB RAM and a RTX 4070 12 GB VRAM usually I am generating with Q8 GGUF checkpoints, no matter if Flux, HIdream or videocheckpoints. Gentime for 33 frames with the loe res was 546 secs including loading checkpoints initially.

1280x720res gentime is 24:01 mins!! Without speedup like fusionX-recipe guess which company gets extremely good vibes....

For GGUF checkpoints thx a lot to the QuantStack community, Checkpoints are here: https://huggingface.co/QuantStack/Wan2.2-T2V-A14B-GGUF

AI-generated commentary

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

Purpose

This Wan2.2 GGUF workflow is for generating video with the Wan Video 14B t2v base model; it takes Wan2.2 GGUF checkpoints and produces video frames.

Inputs and outputs

The named Wan2.2 checkpoint files are Wan2.2-T2V-A14B-HighNoise-Q8_0.gguf and Wan2.2-T2V-A14B-LowNoise-Q8_0.gguf.
The description references a 33-frame low-resolution example and a 1280x720 example.

Setup

It is listed as a Workflows entry published by sikasolutionsworldwide709.
Open the workflow page at https://civitai.com/models/1818827?modelVersionId=2058273.

Dependencies

The workflow lists these node packs: comfyui-gguf, ComfyUI-HunyuanVideoMultiLora, comfyui-kjnodes, comfyui-multigpu, and comfyui-videohelpersuite.
Other listed files are: clip_vision_h.safetensors, umt5_xxl_fp8_e4m3fn_scaled.safetensors, wan_2.1_vae.safetensors, and Wan_bounceV_01.safetensors.
Additional listed model and LoRA files are: wan2.1-i2v-14b-480p-Q8_0.gguf, Wan2.2-T2V-A14B-HighNoise-Q8_0.gguf, Wan2.2-T2V-A14B-LowNoise-Q8_0.gguf, and Wan21_CausVid_14B_T2V_lora_rank32.safetensors.

Configuration

The described setup used Q8 GGUF checkpoints on a PC with 64 GB RAM and an RTX 4070 with 12 GB VRAM.
The description reports 546 seconds for 33 low-resolution frames, including initial checkpoint loading, and 24:01 minutes at 1280x720; treat these as source-reported examples, not tested runtime results.
Suggestion · not verified

Cautions

Before use, verify which exact comfyui-gguf package this workflow expects.
Suggestion · not verified
Before use, verify that your setup has these files: wan_2.1_vae.safetensors, clip_vision_h.safetensors, umt5_xxl_fp8_e4m3fn_scaled.safetensors, Wan_bounceV_01.safetensors, and Wan21_CausVid_14B_T2V_lora_rank32.safetensors.
Suggestion · not verified

Does this need to be edited?

Sources

1 source

CivitaiSourceVersion 2058273

civitai.com/models/1818827

Verified Aug 7

Source excerpts

2 excerpts

NoteAug 7civitai.com/models/1818827

Source context: 2122 downloads · Type Workflows · Base model Wan Video 14B t2v

NoteJul 19civitai.com/models/1818827