Skip to main content
The LTXVAddGuide node adds video conditioning guidance to latent sequences by encoding input images or videos and incorporating them as keyframes into the conditioning data. It processes the input through a VAE encoder and strategically places the resulting latents at specified frame positions while updating both positive and negative conditioning with keyframe information. The node handles frame alignment constraints and allows control over the strength of the conditioning influence.

Inputs

ParameterDescriptionData TypeRequiredRange
positivePositive conditioning input to be modified with keyframe guidanceCONDITIONINGYes-
negativeNegative conditioning input to be modified with keyframe guidanceCONDITIONINGYes-
vaeVAE model used for encoding the input image/video framesVAEYes-
latentInput latent sequence that will receive the conditioning framesLATENTYes-
imageImage or video to condition the latent video on. Must be 8n + 1 frames. If the video is not 8n + 1 frames, it will be cropped to the nearest 8*n + 1 frames.IMAGEYes-
frame_idxFrame index to start the conditioning at. For single-frame images or videos with 1-8 frames, any frame_idx value is acceptable. For videos with 9+ frames, frame_idx must be divisible by 8, otherwise it will be rounded down to the nearest multiple of 8. Negative values are counted from the end of the video. (default: 0)INTNo-9999 to 9999
strengthStrength of the conditioning influence, where 1.0 applies full conditioning and 0.0 applies no conditioning (default: 1.0)FLOATNo0.0 to 10.0
attention_maskOptional pixel-space spatial mask. Controls per-region conditioning influence via self-attention, multiplied by strength.MASKNo-
iclora_parametersOptional IC-LoRA parameters from a Get IC-LoRA Parameters node. Used for adjusting guide processing as required by certain IC-LoRAs (e.g., those with a reference_downscale_factor > 1). When chained, each LTXVAddGuide uses only the parameters connected to it.IC_LORA_PARAMETERSNo-
Note: The input image/video must have a frame count following the 8*n + 1 pattern (e.g., 1, 9, 17, 25 frames). If the input exceeds this pattern, it will be automatically cropped to the nearest valid frame count. Note on iclora_parameters: When using IC-LoRA parameters with a reference_downscale_factor greater than 1, the latent spatial dimensions (width and height) must be divisible by that factor. The node will raise an error if this condition is not met.

Outputs

Output NameDescriptionData Type
positivePositive conditioning updated with keyframe guidance informationCONDITIONING
negativeNegative conditioning updated with keyframe guidance informationCONDITIONING
latentLatent sequence with incorporated conditioning frames and updated noise maskLATENT
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 889348768112c6ecc3ef2e724981d3c49d96339b156617725816cf4186a94b7a