floyo logo
Powered by
ThinkDiffusion
floyo logo
Powered by
ThinkDiffusion

Wan2.2 Animate Character

1.8k

Wan2.2 Animate – Performance-Driven Animation Workflow

Bring your animations to life with real performance video capture using Wan2.2 Animate. Whether you’re creating a 2D cartoon, a 3D animation, or simply swapping out a character in a video, this workflow helps you achieve expressive, performance-driven results.


Key Inputs

  • Resolution

    • Default: 1280×720 (720p)

    • Lower resolutions (e.g., 480p) may work but often fail to capture facial performance effectively.

  • Load Image

    • Upload a still image of the character to animate.

    • The composition does not need to match the first frame of your video, though aligning them improves results.

  • Load Video

    • Provide the driving performance video.

    • Use the frame_load_cap setting to control frame count (default: 53 frames).

  • Prompt

    • Enter a description of the subject.

    • Simple prompts usually work well. For complex scenes, include details about the background, action, or style.

  • Optional Refiner

    • Enable this group to process with the Wan2.2 Low-Noise model.

    • Produces slightly sharper detail at the cost of extra generation time.

    • In many cases, the base generation is sufficient.

Read more

N
Generates in about 8 mins 36 secs

Nodes & Models

GetNode
UNETLoader
Wan2_2-T2V-A14B-LOW_fp8_e4m3fn_scaled_KJ.safetensors
CLIPVisionLoader
clip_vision_h.safetensors
CLIPLoader
umt5_xxl_fp8_e4m3fn_scaled.safetensors
VAELoader
Wan2_1_VAE_bf16.safetensors
Label (rgthree)
Fast Groups Bypasser (rgthree)
LoadImage
CLIPTextEncode
SetNode
Reroute
WanVideoTorchCompileSettings
WanVideoBlockSwap
WanVideoLoraSelectMulti
WanAnimate_relight_lora_fp16.safetensors
lightx2v_I2V_14B_480p_cfg_step_distill_rank128_bf16.safetensors
WanVideoVAELoader
Wan2_1_VAE_bf16.safetensors
WanVideoModelLoader
Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors
WanVideoSetLoRAs
WanVideoTorchCompileSettings
WanVideoBlockSwap
WanVideoLoraSelectMulti
WanAnimate_relight_lora_fp16.safetensors
lightx2v_I2V_14B_480p_cfg_step_distill_rank128_bf16.safetensors
WanVideoVAELoader
Wan2_1_VAE_bf16.safetensors
WanVideoModelLoader
Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors
WanVideoClipVisionEncode
WanVideoSetLoRAs
WanVideoAnimateEmbeds
WanVideoTextEmbedBridge
WanVideoSetBlockSwap
WanVideoSampler
FaceMaskFromPoseKeypoints
WanVideoDecode
INTConstant
INTConstant
ImageResizeKJv2
ImageCropByMaskAndResize
Text Multiline
VHS_LoadVideo
VHS_VideoCombine
ImageResizeKJv2
ImageCropByMaskAndResize
PixelPerfectResolution
DWPreprocessor
PixelPerfectResolution
PixelPerfectResolution
DWPreprocessor
DWPreprocessor
UltimateSDUpscaleNoUpscale

Wan2.2 Animate – Performance-Driven Animation Workflow

Bring your animations to life with real performance video capture using Wan2.2 Animate. Whether you’re creating a 2D cartoon, a 3D animation, or simply swapping out a character in a video, this workflow helps you achieve expressive, performance-driven results.


Key Inputs

  • Resolution

    • Default: 1280×720 (720p)

    • Lower resolutions (e.g., 480p) may work but often fail to capture facial performance effectively.

  • Load Image

    • Upload a still image of the character to animate.

    • The composition does not need to match the first frame of your video, though aligning them improves results.

  • Load Video

    • Provide the driving performance video.

    • Use the frame_load_cap setting to control frame count (default: 53 frames).

  • Prompt

    • Enter a description of the subject.

    • Simple prompts usually work well. For complex scenes, include details about the background, action, or style.

  • Optional Refiner

    • Enable this group to process with the Wan2.2 Low-Noise model.

    • Produces slightly sharper detail at the cost of extra generation time.

    • In many cases, the base generation is sufficient.

Read more

N