Create Audio To Video AI With Wan 2.2 S2V, Fast Generation For Low VRAM GPU Comfy UI Workflow

Sep 3, 2025

CompareAI.AI

Generate realistic AI videos from a single image and an audio using the latest open-source WAN 2.2 S2V (still-to-video) model in ComfyUI. Grab the ready-to-use workflow plus all required models and resources below and start creating high-quality videos in minutes.

What is ComfyUI? & How to set it up on your system? —> Click Here

Download the workflow below to create AI videos with WAN 2.2 S2V. Direct download links for the S2V model and required resources are provided underneath.

Simple ComfyUI Workflow for WAN 2.2 S2V (Audio + Image-to-Video)

Note : On clicking download button, It may open a json file in a new tab, Just save the json file and open it in comfy UI.

Watch the below youtube video to understand the above ComfyUI workflow.

MODELS USED IN THIS WORKFLOW WITH DOWNLOAD LINKS

⚛️ Load Diffusion Model
Download 2.2 S2V gguf models
[🔗 HuggingFace Link - Click Here]
Place in: `ComfyUI/models/diffusion_models` }

🟣 Load Clip
Download UMT5XXL FP8 (Scaled)
[🔗 Download umt5_xxl_fp8_e4m3fn_scaled.safetensors]
Place in: `ComfyUI/models/clip`

🟣 Load VAE
Download wan_2.1_vae.safetensors
[🔗 Download wan_2.1_vae.safetensors]
Place in: `ComfyUI/models/vae`

🟣 Wav2vec2 Model Loader
Download wav2vec2-chinese-base_fp16.safetensors
[🔗 Download wav2vec2-chinese-base_fp16.safetensors]
Place in: `ComfyUI/models/wav2vec2`
(Create that wav2vec2 folder if you don't have it.)

🟣 Load Lora
Download wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors
[🔗 Download wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors]
Place in: `ComfyUI/models/loras`
(Download the high noise model and rename it as "wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors")

Check out our comparisons on image and video generative AI tools like Leonardo AI, Ideogram etc. with results and analysis over each AI tool, Or just judge them yourself by comparing the results of each ai tool, go to home now!.

Create Audio To Video AI With Wan 2.2 S2V, Fast Generation For Low VRAM GPU Comfy UI Workflow

Sep 3, 2025

CompareAI.AI

Generate realistic AI videos from a single image and an audio using the latest open-source WAN 2.2 S2V (still-to-video) model in ComfyUI. Grab the ready-to-use workflow plus all required models and resources below and start creating high-quality videos in minutes.

What is ComfyUI? & How to set it up on your system? —> Click Here

Download the workflow below to create AI videos with WAN 2.2 S2V. Direct download links for the S2V model and required resources are provided underneath.

Simple ComfyUI Workflow for WAN 2.2 S2V (Audio + Image-to-Video)

Note : On clicking download button, It may open a json file in a new tab, Just save the json file and open it in comfy UI.

Watch the below youtube video to understand the above ComfyUI workflow.

MODELS USED IN THIS WORKFLOW WITH DOWNLOAD LINKS

⚛️ Load Diffusion Model
Download 2.2 S2V gguf models
[🔗 HuggingFace Link - Click Here]
Place in: `ComfyUI/models/diffusion_models` }

🟣 Load Clip
Download UMT5XXL FP8 (Scaled)
[🔗 Download umt5_xxl_fp8_e4m3fn_scaled.safetensors]
Place in: `ComfyUI/models/clip`

🟣 Load VAE
Download wan_2.1_vae.safetensors
[🔗 Download wan_2.1_vae.safetensors]
Place in: `ComfyUI/models/vae`

🟣 Wav2vec2 Model Loader
Download wav2vec2-chinese-base_fp16.safetensors
[🔗 Download wav2vec2-chinese-base_fp16.safetensors]
Place in: `ComfyUI/models/wav2vec2`
(Create that wav2vec2 folder if you don't have it.)

🟣 Load Lora
Download wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors
[🔗 Download wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors]
Place in: `ComfyUI/models/loras`
(Download the high noise model and rename it as "wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors")

Check out our comparisons on image and video generative AI tools like Leonardo AI, Ideogram etc. with results and analysis over each AI tool, Or just judge them yourself by comparing the results of each ai tool, go to home now!.

Create Audio To Video AI With Wan 2.2 S2V, Fast Generation For Low VRAM GPU Comfy UI Workflow

Sep 3, 2025

CompareAI.AI

Generate realistic AI videos from a single image and an audio using the latest open-source WAN 2.2 S2V (still-to-video) model in ComfyUI. Grab the ready-to-use workflow plus all required models and resources below and start creating high-quality videos in minutes.

What is ComfyUI? & How to set it up on your system? —> Click Here

Download the workflow below to create AI videos with WAN 2.2 S2V. Direct download links for the S2V model and required resources are provided underneath.

Simple ComfyUI Workflow for WAN 2.2 S2V (Audio + Image-to-Video)

Note : On clicking download button, It may open a json file in a new tab, Just save the json file and open it in comfy UI.

Watch the below youtube video to understand the above ComfyUI workflow.

MODELS USED IN THIS WORKFLOW WITH DOWNLOAD LINKS

⚛️ Load Diffusion Model
Download 2.2 S2V gguf models
[🔗 HuggingFace Link - Click Here]
Place in: `ComfyUI/models/diffusion_models` }

🟣 Load Clip
Download UMT5XXL FP8 (Scaled)
[🔗 Download umt5_xxl_fp8_e4m3fn_scaled.safetensors]
Place in: `ComfyUI/models/clip`

🟣 Load VAE
Download wan_2.1_vae.safetensors
[🔗 Download wan_2.1_vae.safetensors]
Place in: `ComfyUI/models/vae`

🟣 Wav2vec2 Model Loader
Download wav2vec2-chinese-base_fp16.safetensors
[🔗 Download wav2vec2-chinese-base_fp16.safetensors]
Place in: `ComfyUI/models/wav2vec2`
(Create that wav2vec2 folder if you don't have it.)

🟣 Load Lora
Download wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors
[🔗 Download wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors]
Place in: `ComfyUI/models/loras`
(Download the high noise model and rename it as "wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors")

Check out our comparisons on image and video generative AI tools like Leonardo AI, Ideogram etc. with results and analysis over each AI tool, Or just judge them yourself by comparing the results of each ai tool, go to home now!.