Skip to content
BytePlusBytePlusBytePlus ModelArk for ComfyUI

Text to Image to Video

Template
Text to Image to Video, in the ComfyUI template browser under BytePlus
Workflow file
Text to Image to Video.json
BytePlus nodes
Seedream 4.5 & 5.0, Seedance 2.5 First-Last-Frame to Video
Keys needed
ModelArk API key
Also uses
Save Image, Save Video (built into ComfyUI)

What it makes

Section titled “What it makes”

A still image and a short video from one idea. Seedream paints an astronaut on a red desert dune with a huge ringed planet in the sky. Seedance then uses that picture as the first frame and animates it, so the clip starts exactly where the image is. Writing the image first gives you control over the look before you pay for the video.

The workflow

Section titled “The workflow”
Seedream connected to Save First Frame and to the first_frame input of Seedance First-Last-Frame to Video, which feeds Save Video
Four nodes: Seedream, Save First Frame, Seedance and Save Video.
  1. Seedream 4.5 & 5.0 makes the image from the prompt “A lone astronaut standing on a red desert dune at sunrise, a huge ringed planet filling the sky, cinematic wide shot…”. The model is seedream 5.0 pro and the size is 2048x1152 (16:9). No reference image is connected.
  2. Save First Frame keeps the image in the output folder.
  3. Seedance 2.5 First-Last-Frame to Video gets the image as first_frame and animates it with the model Seedance 2.0 at 720p, ratio adaptive, 5 seconds, with audio. Its prompt says the astronaut turns toward the camera while dust drifts and the camera pushes in. last_frame stays empty.
  4. Save Video keeps the clip.

Run it

Section titled “Run it”
  1. Add your ModelArk API key in Settings → BytePlus if you have not yet (API keys).
  2. Open the template browser and choose Text to Image to Video under BytePlus.
  3. Press Run. The image appears first in Save First Frame, then the video in Save Video when Seedance finishes.

Settings worth changing

Section titled “Settings worth changing”
  • Seedream prompt: describe the scene you want as the first frame. Write the Seedance prompt about the motion, not about the scene again.
  • Seedream size_preset: the video ratio is adaptive, so the clip follows the shape of the image. Pick another preset for a vertical or square clip.
  • Seedance model: Seedance 2.0 is set. The list also has Seedance 2.5 and the Draft, Premium, Fast and Mini variants. Seedance 2.5 models have no ratio input.
  • Seedance resolution and duration: 720p and 5 seconds are set. Raise them when you are happy with the motion.
  • last_frame: connect a second image to make the clip end on a given picture.

Example result

Section titled “Example result”
The first frame of the astronaut on the dune, and three frames of the video below it
Example result: the first frame and the video.
The astronaut clip made from the Seedream image.

Time and cost

Section titled “Time and cost”

ModelArk bills the image once and the video task once. The seed controls are set to randomize, so every run makes new results and bills again. Set a control to fixed and ComfyUI reuses the cached result of an unchanged node without billing.

Section titled “Related”