Nodes/comfyui-budgetpixel/SeeDance 1.5 Pro (BudgetPixel)
ComfyUI Node

SeeDance 1.5 Pro (BudgetPixel)

SeeDance 1.5 Pro — the cheap way to animate a still

By budgetpixel-ai·Created about 17 hours ago·Updated about 12 hours ago· 3
SeeDance 1.5 Pro (BudgetPixel)
  • end_image
  • image
  • video
prompt
aspect_ratio16:9
generate_audiofalse
length_seconds5
resolution720p
seed-1
force_rerunfalse

SeeDance 1.5 Pro is ByteDance's video model at its most approachable - and at 60 credits per second, it's the cheapest Seedance in the pack. That combination makes it the node you reach for when you want to learn the video side of the BudgetPixel pack without watching your credit balance evaporate. Text-to-video works, but the real use is image-to-video: feed it a start frame, optionally an end frame, and the model interpolates the motion between them. Two stills and a prompt is the classic animatic trick, and it's where this model shines.

What it does

It's a hosted API node - no local weights, no GPU, the generation runs on BudgetPixel's servers and the finished clip comes back into your graph as a core VIDEO value. Length is a fixed set of choices (5, 10, or 12 seconds) rather than a free-form slider, and resolution runs 480p/720p/1080p. Optionally flip generate_audio on to get sound with the clip.

The inputs that matter

  • prompt - required, describes the video (or the motion, for image-to-video).
  • image - the start frame. Wire any IMAGE into it.
  • end_image - optional last frame. Used with image, it defines the interpolation: the model generates the video between frame A and frame B. This is the underrated input.
  • length_seconds - 5/10/12. Billing is per second, so 12s is 2.4× the 5s cost.
  • resolution - 480p/720p/1080p.
  • generate_audio - off by default; flip it on if the clip wants sound.
  • seed - -1 random; set a value to reproduce.

Output is a single video (VIDEO) - wire it into a Save Video node or a frame extractor. You'll need a recent ComfyUI (2025+) for the core video type to exist at all.

How it works

The standard pack flow: image inputs are serialized and uploaded, a video job is created, the node polls every few seconds (video jobs are slow - minutes, not seconds) and then downloads the MP4 into the temp folder and wraps it as a VIDEO value. Progress shows in the bar; don't assume it's hung.

Install and key setup

ComfyUI Manager → search BudgetPixel → install, or:

cd ComfyUI/custom_nodes
git clone https://github.com/budgetpixel-ai/comfyui-budgetpixel.git
pip install -r comfyui-budgetpixel/requirements.txt

Only requests as a dependency. API key from a paid plan via BUDGETPIXEL_API_KEY or budgetpixel.json; keys never land in workflow JSON.

Where people get burned

Video billing is the sharp edge here - per second, every time, and an interrupted job still finishes and bills server-side. A 12-second 1080p clip at 60 credits/second is 720 credits; at 5 seconds it's 300. Draft at 480p and short. Also remember ComfyUI's caching: unchanged inputs won't re-fire a generation, so use force_rerun when you want a fresh take - and a fresh bill.

CategoryBudgetPixel/Video

Inputs (9)

NameTypeDefaultDescription
promptSTRINGText description of the video.
aspect_ratiooptCOMBO16:9Aspect ratio.
end_imageoptIMAGEOptional end frame, used together with `image` (the start frame) to interpolate the video between the two frames. Same input forms as `image`.
generate_audiooptBOOLEANfalseGenerate audio with the video.
imageoptIMAGEOptional start frame for image-to-video. Provide a public image URL, a data URI, raw base64, or an uploaded-file URL from POST /v1/uploads. Omit for text-to-video.
length_secondsoptCOMBO5Video length in seconds.
resolutionoptCOMBO720pOutput resolution.
seedoptINT-1-1–2147483647Seed for reproducible generation. Omit for random.
force_rerunoptBOOLEANfalse

Outputs (1)

NameTypeDescription
videoVIDEO