floyo logo
Powered by
ThinkDiffusion
⚡️Nano Banana 2 ⚡️ just landed. Start creating now.
floyo logo
Powered by
ThinkDiffusion
⚡️Nano Banana 2 ⚡️ just landed. Start creating now.

Image to Video with Wan

Turn still images into amazing videos with just prompts, using the SOTA Wan video model.

941

Generates in about 7 mins 21 secs

Nodes & Models

Label (rgthree)
Note
CLIPVisionLoader
clip_vision_h.safetensors
LoadImage
CLIPLoader
umt5_xxl_fp8_e4m3fn_scaled.safetensors
VAELoader
wan_2.1_vae.safetensors
UNETLoader
Wan2_1-I2V-14B-480P_fp8_e4m3fn.safetensors
CLIPVisionEncode
CLIPTextEncode
WanImageToVideo
ModelSamplingSD3
KSampler
VAEDecode
ImageResizeKJ
ImageResizeKJ
ColorMatch
ImageResizeKJ
ColorMatch
TeaCache
VHS_VideoCombine

Turn still images into amazing videos with just prompts, using the SOTA Wan video model.

Key Inputs

Load: Use any JPG or PNG showing your subject clearly

Width & height: Make sure the resolution does not exceed 1280 x 720. Once exceeded, there will be a decrease in quality.

Keep Proportion: Enable keep_proportion if you want to keep the same size for the input and output.

Prompt: as descriptive a prompt as possible

Video Length: 96 Frames maximum

File Format: H.264 and more.

Output Path

../outputs/Image2Video_Wan/

Workflow

Read more

N