Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.comfy.org/llms.txt

Use this file to discover all available pages before exploring further.

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Topaz Video Enhance V2

The Topaz Video Enhance V2 node allows you to upscale and enhance video using Topaz Labs’ AI models. It can increase resolution, adjust frame rate through interpolation, and apply creative or realistic enhancements to breathe new life into your video footage.

Inputs

ParameterData TypeRequiredRangeDescription
videoVIDEOYes-The input video to be processed. Must be in MP4 container format.
upscaler_modelCOMBOYes"Astra 2"
"Starlight (Astra) Fast"
"Starlight (Astra) Creative"
"Starlight Precise 2.5"
"Disabled"
The AI model used for upscaling the video. Selecting “Disabled” means no upscaling will be applied.
upscaler_model.upscaler_resolutionCOMBOConditional"FullHD (1080p)"
"4K (2160p)"
The target output resolution for the upscaler. This parameter is required when an upscaler model is selected (not “Disabled”).
upscaler_model.creativityFLOAT / COMBOConditionalAstra 2: 0.0 to 1.0 (step 0.1)
Starlight Creative: "low"
"middle"
"high"
Creative strength of the upscale. Available only for “Astra 2” and “Starlight (Astra) Creative” models. For Astra 2, it’s a slider (default: 0.5). For Starlight Creative, it’s a combo (default: “low”).
upscaler_model.promptSTRINGNo-Optional descriptive (not instructive) scene prompt. Only available for the “Astra 2” model. Capped at 450 input frames (~15s @ 30fps) when set. Default: empty.
upscaler_model.sharpFLOATNo0.0 to 1.0 (step 0.01)Pre-enhance sharpness: 0.0=Gaussian blur, 0.5=passthrough (default), 1.0=USM sharpening. Only available for the “Astra 2” model. Default: 0.5.
upscaler_model.realismFLOATNo0.0 to 1.0 (step 0.01)Pulls output toward photographic realism. Leave at 0 for the model default. Only available for the “Astra 2” model. Default: 0.0.
interpolation_modelCOMBOYes"Disabled"
"apo-8"
The AI model used for frame interpolation. Selecting “Disabled” means no interpolation will be applied.
interpolation_model.interpolation_frame_rateINTConditional15 to 240Output frame rate. Required when interpolation model is “apo-8”. Default: 60.
interpolation_model.interpolation_slowmoINTNo1 to 16Slow-motion factor applied to the input video. For example, 2 makes the output twice as slow and doubles the duration. Default: 1.
interpolation_model.interpolation_duplicateBOOLEANNoTrue/FalseAnalyze the input for duplicate frames and remove them. Default: False.
interpolation_model.interpolation_duplicate_thresholdFLOATNo0.001 to 0.1 (step 0.001)Detection sensitivity for duplicate frames. Default: 0.01.
dynamic_compression_levelCOMBONo"Low"
"Mid"
"High"
CQP level for video compression. Default: “Low”.
Important Constraints:
  • At least one of upscaler_model or interpolation_model must be enabled (not “Disabled”), otherwise an error is raised.
  • The input video must be in MP4 container format.
  • The “Astra 2” model with a prompt is limited to 450 input frames (~15 seconds at 30fps). Without a prompt, it is limited to 9000 input frames.
  • When upscaler_model is not “Disabled”, the upscaler_resolution sub-parameter is required.
  • When interpolation_model is not “Disabled”, the interpolation_frame_rate sub-parameter is required.

Outputs

Output NameData TypeDescription
videoVIDEOThe enhanced video output after applying the selected upscaling and/or interpolation filters.

Source fingerprint (SHA-256): 3b3df144ecd7de6e0c141cffa30fc7b303e0edc991e714c2f0aaf0e4be903166