Wan2.2 Sound-to-Video in ComfyUI: Create Audio-Reactive AI Videos Locally

About This Workflow

Generate AI videos driven by audio input, where visuals react to music, voice, or sound effects.

Create beat-synced and rhythm-aware animations that follow tempo, volume changes, and audio intensity.

Produce music videos, visualizers, reels, and short clips without manually animating frames.

Control motion strength, video length, frame rate, and visual style directly inside ComfyUI.

Run the entire sound-to-video workflow locally on your GPU, ensuring privacy, unlimited usage, and no cloud costs.

Required Models

Model links

Diffusion Model

• wan2.2_s2v_14B_bf16.safetensors
• wan2.2_s2v_14B_fp8_scaled.safetensors

VAE

• wan_2.1_vae.safetensors

Text Encoder

• umt5_xxl_fp8_e4m3fn_scaled.safetensors

Lora

• Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank256_bf16.safetensors

Audio Encoders

• wav2vec2_large_english_fp16.safetensors

File save location

ComfyUI/
├───📂 models/
│   ├───📂 diffusion_models/
│   │   ├─── wan2.2_i2v_low_noise_14B_fp16.safetensors
│   │   └─── wan2.2_i2v_high_noise_14B_fp16.safetensors
│   ├───📂 text_encoders/
│   │   └─── umt5_xxl_fp8_e4m3fn_scaled.safetensors 
│   └───📂 vae/
│       └── wan_2.1_vae.safetensors
│   ├───📂 audio_encoders/
│   │   └─── wav2vec2_large_english_fp16.safetensors 
│   └───📂 lora/
│       └── lightx2v_I2V_14B_480p_cfg_step_distill_rank256_bf16.safetensors

Workflow Graph