Nodes/Comfyui-zhenzhen/zhenzhen-seedance2-low-price
ComfyUI Node

zhenzhen-seedance2-low-price

The domestic low-price video workhorse

By T8mars·Created about a year ago·Updated 2 days ago· 740
zhenzhen-seedance2-low-price
  • api_config
  • first_image
  • last_image
  • image1
  • image2
  • image3
  • image4
  • image5
  • image6
  • image7
  • image8
  • image9
  • video1
  • video2
  • video3
  • audio1
  • audio2
  • audio3
  • video
  • video_url
  • task_id
  • response
modetext_to_video
regiondomestic
tiermini
prompt
seconds5
resolution480p
ratioadaptive
generate_audiotrue
return_last_framefalse
seed-1
skip_errorfalse

This is the flagship of zhenzhen's domestic low-price line, and the README is proud of it: Seedance 2.0 through the api.seedance.nz shop at roughly 75% of the overseas price. Comfly_seedance2_low_price does text-to-video, image-to-video, and full multimodal video (images + clips + audio as references), across three tiers and resolutions from 480p all the way to native 4K. If you want Seedance video in ComfyUI and you're price-sensitive, this is the node you'd reach for first.

Like every low-price node, it takes an api_config from the pack's Comfly_seedance2_low_price_settings node rather than its own api_key field. That settings node carries the base URL (https://api.seedance.nz by default) and your token - one settings node feeds the whole domestic family.

How it works

Async submit → poll → download. Your reference media is uploaded to the shop first, the job runs server-side, and the node returns a real VIDEO tensor plus video_url, task_id, and response. The mode + region + tier trio is the interesting part:

  • mode - text_to_video, image_to_video (first + optional last frame), or multimodal_video (up to 9 images, 3 videos, 3 audio).
  • region - domestic or global. Domestic is the cheap route; global routes overseas.
  • tier - mini (the token-priced half-price tier), fast, or standard.

The inputs that matter

  • prompt - required.
  • seconds - 4–15 (the -1 option means "auto/longest").
  • resolution - 480p/720p/1080p/2k/4k, plus the two native options (native1080p, native4k) that the README calls out for their filmic quality - "效果炸裂," per the changelog. Native tiers can be slow and don't queue like the standard ones.
  • ratio - adaptive (default) through 21:9.
  • generate_audio - on by default.
  • return_last_frame - handy for chaining.
  • first_image / last_image - the image-to-video pair.
  • image1image9, video1video3, audio1audio3 - multimodal references. The node validates which slots fit which mode and errors helpfully if you mix them (t2v rejects all references; i2v only accepts first/last).

seed is ComfyUI cache-control only and isn't sent to the API.

Installing it

cd ComfyUI/custom_nodes
git clone https://github.com/T8mars/Comfyui-zhenzhen

or ComfyUI Manager (search "Comfyui-zhenzhen"), restart, wire a settings node with your api.seedance.nz token. No model downloads.

Common issues

  • "text_to_video does not accept reference media" - you left images in the slots while in t2v mode. Clean the slots or switch modes.
  • Domestic fast tier audio cap - the node refuses audio uploads over ~15MB on the domestic fast tier. Trim your reference clips.
  • Timeouts on 4K / native tiers - these take a long time. If the node times out, the job may still be finishing; grab the task_id and check the shop's async-task page before paying again.

For anyone on the domestic shop, this is the core Seedance node - start here, then branch to 2.5 (Comfly_seedance25_standard_low_price) or the FAL multi-ref variant (Comfly_seedance2_fal) when the job demands it.

Categoryzhenzhen/Seedance2 Low Price

Inputs (29)

NameTypeDefaultDescription
modeCOMBOtext_to_video3 options: text_to_video, image_to_video, multimodal_video
regionCOMBOdomestic2 options: domestic, global
tierCOMBOmini3 options: mini, fast, standard
promptSTRING
secondsCOMBO513 options: -1, 4, 5, 6, 7, 8, +7
resolutionCOMBO480p7 options: 480p, 720p, 1080p, 2k, 4k, native1080p, +1
ratioCOMBOadaptive7 options: adaptive, 16:9, 4:3, 1:1, 3:4, 9:16, +1
generate_audiooptBOOLEANtrue
return_last_frameoptBOOLEANfalse
seedoptINT-1-1–2147483647
api_configoptZHENZHEN_SEEDANCE2_CONFIG
first_imageoptIMAGE
last_imageoptIMAGE
image1optIMAGE
image2optIMAGE
image3optIMAGE
image4optIMAGE
image5optIMAGE
image6optIMAGE
image7optIMAGE
image8optIMAGE
image9optIMAGE
video1optVIDEO
video2optVIDEO
video3optVIDEO
audio1optAUDIO
audio2optAUDIO
audio3optAUDIO
skip_erroroptBOOLEANfalse

Outputs (4)

NameTypeDescription
videoVIDEO
video_urlSTRING
task_idSTRING
responseSTRING