ComfyUI Node

Venice AI · Video

Generate videos with any Venice AI video model (needs your own Venice API key).

By Slartibart23·Created 25 days ago·Updated 24 days ago· 1
Venice AI · Video
  • image
  • end_image
  • reference_images
  • video
  • audio
  • video
  • video_path
  • last_frame
  • prompt
  • info
◄model▾►
◄prompt►
◄negative_prompt►
◄duration5s►
◄resolution720p►
◄aspect_ratioauto►
◄generate_audiotrue►
◄previewfalse►
◄price_onlyfalse►
◄max_price_usd2.00►
◄seedance_consentfalse►
◄keep_raw_copytrue►
◄timeout_minutes30►
◄variation0►
◄api_key►
CategoryVenice AI

Inputs (20)

NameTypeDefaultDescription
modelCOMBO★ = your favourite (Venice_AI_favourites.txt in the ComfyUI folder). T2V/I2V/R2V/V2V = text/image/reference/video to video. Price at about 5 s 720p. UNCENSORED/censored as flagged by Venice; 'private' = Venice does not pass your data on.
promptSTRING—
negative_promptSTRING—
durationCOMBO5sThe list shows what the selected model can do, up to its maximum.
resolutionCOMBO720p15 options: 2x, 4x, 360p, 480P, 480p, 540p, +9
aspect_ratioCOMBOautoauto = the model's own auto/adaptive mode, else the ratio closest to the input image (16:9 without image).
generate_audioBOOLEANtrueOnly for models with switchable audio.
previewBOOLEANfalseCheap test run: lowest resolution and ~4-5 s; models that cannot go below 1080p use their cheaper sibling. Not the same video as the full run.
price_onlyBOOLEANfalseOnly ask Venice for the price, generate nothing.
max_price_usdFLOAT2.000–100Stops before anything is billed when the quoted price is higher.
seedance_consentBOOLEANfalseSeedance only: you confirm the provider terms, that you have the legal right to use the input media (incl. faces) and that it may be screened.
keep_raw_copyBOOLEANtrueKeep the downloaded MP4 in output/venice_ai (paid results are never lost). Off = ComfyUI temp folder, deleted on restart.
timeout_minutesINT301–240—
variationINT00–999999999Venice video has no seed. Change this (or set it to randomize) to get a new take with identical settings.
imageoptIMAGEStart image (I2V). For R2V it is used as first reference.
end_imageoptIMAGEEnd frame, for models with first/last frame support.
reference_imagesoptIMAGER2V: up to 30 reference images (image batch).
videooptVIDEOV2V: input video for edit / recast / extend / upscale.
audiooptAUDIOFor models with audio input (max. 30 s).
api_keyoptSTRINGYour Venice API key. NOT recommended here: it is saved inside workflows and every saved file. Prefer Venice_AI_API.txt in the ComfyUI folder or the VENICE_API_KEY variable.

Outputs (5)

NameTypeDescription
videoVIDEO—
video_pathSTRING—
last_frameIMAGE—
promptSTRING—
infoSTRING—