LTX 2.3 - Motion Transfer
Copy any video's motion onto a still character photo using LTX 2.3
image-to-video
LTX 2.3
LTXV
motion transfer
video generation
video-to-video
4
479
Nodes & Models
LoadImage
PrimitiveInt
VAELoaderKJ
LTX23_audio_vae_bf16.safetensors
LTX23_video_vae_bf16.safetensors
UNETLoader
ltx-2.3-22b-distilled_transformer_only_fp8_input_scaled_v3.safetensors
DualCLIPLoader
gemma_3_12B_it_fp8_scaled.safetensors
ltx-2.3_text_projection_bf16.safetensors
KSamplerSelect
ManualSigmas
RandomNoise
LTXICLoRALoaderModelOnly
ltx-2.3-22b-ic-lora-union-control-ref0.5-ltx-2-3-ic-lora-control-MplWpT-2.safetensors
LoraLoader
gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors
ComfyMathExpression
ComfySwitchNode
LTXVEmptyLatentAudio
CLIPTextEncode
LTXVConditioning
SamplerCustomAdvanced
LTXVPreprocess
EmptyLTXVLatentVideo
LTXVAudioVAEDecode
LTXVSeparateAVLatent
CFGGuider
LTXVConcatAVLatent
VAEDecodeTiled
GetImageSize
ImageResizeKJv2
LTXAddVideoICLoRAGuide
AIO_Preprocessor
LTXVCropGuides
LTXVImgToVideoInplace
GetImageRangeFromBatch
PreviewImage
easy compare
ImageFromBatch
ImageStitch
VHS_LoadVideo
VHS_VideoCombine
ABOUT THE WORKFLOW
Transfer Motion onto a Character
Upload an image of the character you want to animate and a video showing the motion you want them to perform. Write a short prompt describing the scene. The workflow reads the motion from your reference video and applies it to your character, with LTX 2.3 rendering the final clip and its synced audio.
HOW IT WORKS
Step 1. Upload your character image
A clear photo of the character you want animated.
Step 2. Upload your motion reference video
A clip showing the action you want copied over. The model reads body movement and timing from this video.
Step 3. Write a short prompt
Describe the scene, e.g. "a panda is dancing." Keep it simple since there's no auto-enhancer to fill in gaps.
Step 4. Hit run
The workflow generates a new video of your character performing the reference motion, with audio.
RECOMMENDED SETTINGS
Length: Follows the 8N+1 rule. 121 frames = 5 sec, 201 = 8 sec, 249 = 10 sec, 297 = 12 sec.
Width/Height: Default 1280 x 720. Keep as a multiple of 32.
WHAT WORKS BEST / WHAT TO AVOID
✅ Clear character images, smooth motion reference clips, prompts that match the action shown
⚠️ Blurry character photos, fast or chaotic motion, prompts that contradict the reference video
Read more






_1783698095782.gif?width=400&height=300&quality=80&resize=cover)





