Wan 2.2 GGUF T2V 14B Workflow for ComfyUI (Beginner Friendly)

This ComfyUI workflow is for text-to-video generation: it takes text conditioning and produces a combined video using Wan Video 2.2 T2V-A14B.

The graph encodes text, sets up video latents, samples with GGUF models, decodes with a VAE, and combines the video.

Enter text conditioning through CLIPTextEncode.

The output path uses VHS_VideoCombine to combine the video.

A usage video is provided with the workflow.

The workflow is published on Civitai by Lordcaocao2025.

The listed model files are: lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank128_bf16.safetensors, umt5_xxl_fp16.safetensors, wan_2.1_vae.safetensors, Wan2.2-T2V-A14B-HighNoise-Q5_K_S.gguf, and Wan2.2-T2V-A14B-LowNoise-Q5_K_S.gguf.

The listed node packs are comfyui-gguf and comfyui-videohelpersuite.

Before use, confirm which package provides comfyui-gguf.

Vorschlag · nicht geprüft

Review KSamplerAdvanced, ModelSamplingSD3, and LoraLoader for sampling, model-sampling, and LoRA settings.

Vorschlag · nicht geprüft

The graph uses UnetLoaderGGUF with separate HighNoise and LowNoise Wan2.2 model files.

Before use, verify that you have the exact files named Wan2.2-T2V-A14B-HighNoise-Q5_K_S.gguf, Wan2.2-T2V-A14B-LowNoise-Q5_K_S.gguf, and wan_2.1_vae.safetensors.

Vorschlag · nicht geprüft

No model-use license is specified; review applicable terms before use.

Vorschlag · nicht geprüft