Kandinsky 5.0 I2V Lite

Note: All model files and custom nodes below must be downloaded via the client. If you do not have the client, Download >>>

A lightweight 2B parameter model that ranks among the top open-source video generation models, capable of generating videos up to 10 seconds long.

Hunyuan Video VAE >2 files

Name	Size	Target Dir
hunyuan_video_vae_fp8_e4m3fn.safetensors	247 MB	vae
hunyuan_video_vae_bf16.safetensors	493 MB	vae

clip_l >1 files

Name	Size	Target Dir
clip_l.safetensors	246 MB	text_encoders

Qwen2.5 VL 7B >2 files

Name	Size	Target Dir
qwen_2.5_vl_7b_fp8_scaled.safetensors	9.38 GB	text_encoders
qwen_2.5_vl_7b.safetensors	16.58 GB	text_encoders

Kandinsky 5.0 I2V Lite >1 files

Name	Size	Target Dir
kandinsky5lite_i2v_5s.safetensors	4.57 GB	diffusion_models