Concurrent Submit | zhenzhen-seedance2-low-price
Tiers, regions and start-end frames
- api_config
- first_image
- last_image
- image1
- image2
- image3
- image4
- image5
- image6
- image7
- image8
- image9
- video1
- video2
- video3
- audio1
- audio2
- audio3
- task
Seedance 2.0, ByteDance's video model, on the domestic discount site - the node that started the whole "low-price" family back in July 2026. It's the most configurable Seedance 2.0 wrapper in the pack: three modes, three tiers, two regions, up to 4K output, and both start/end-frame and multimodal reference inputs.
The headline structure: mode (text_to_video / image_to_video / multimodal_video), tier (mini / fast / standard - mini is the budget lane at roughly half the standard price, per the changelog), and region (domestic / global). Together those three dropdowns pick your price point, your quality bar, and your routing. native4k and native1080p resolution options are where the "电影级" (cinematic) claims come from - and where the bill balloons.
Like every node in the low-price family it takes an api_config from the Comfly Seedance2 Low Price Settings node (Comfly_seedance2_low_price_settings, which holds base_url + api_key) instead of a bare key field. It's a video node: shared video pool (10 workers), COMFLY_VIDEO_FUTURE output, collected by Concurrent Collect Videos (ComflyConcurrent_Video_Await) into video_1…video_10.
The inputs
- prompt, seconds (-1 or 4–15), resolution (480p→4k, native1080p, native4k), ratio - the core generation set.
- first_image / last_image - the start and end frames for that classic interpolation look.
last_imageis the killer feature people buy Seedance 2.0 for. - image1…image9, video1…video3, audio1…audio3 - multimodal references for the
multimodal_videomode. - generate_audio - on by default (synchronized sound); return_last_frame - off.
- seed - -1 (random) to 2147483647.
Install
cd ComfyUI/custom_nodes
git clone https://github.com/T8mars/Comfyui-zhenzhen
Or ComfyUI Manager → search Comfyui-zhenzhen → restart, then build api_config with the settings node and your domestic-site token (api.seedance.nz).
Gotchas
- Tier = budget, and mini is genuinely cheaper. The changelog puts mini at ~half the token price of standard at 720p. For drafts and tests, mini is the right default.
- Reference video seconds are billed. Multimodal mode counts your reference clips toward the total; the README has said this since day one.
- Region matters outside China.
domesticis optimized for the mainland site; if you're elsewhere,globalplus the overseas base_url is the reliable combo. The standing pack advice: connection errors are usually your route to their server, not your node. - The 1-hour poll ceiling on big jobs. native4k at 15s can run long - if it times out, pull the task from the async-task page before resubmitting.
Inputs (29)
| Name | Type | Default | Description |
|---|---|---|---|
| mode | COMBO | text_to_video | 3 options: text_to_video, image_to_video, multimodal_video |
| region | COMBO | domestic | 2 options: domestic, global |
| tier | COMBO | mini | 3 options: mini, fast, standard |
| prompt | STRING | — | |
| seconds | COMBO | 5 | 13 options: -1, 4, 5, 6, 7, 8, +7 |
| resolution | COMBO | 480p | 7 options: 480p, 720p, 1080p, 2k, 4k, native1080p, +1 |
| ratio | COMBO | adaptive | 7 options: adaptive, 16:9, 4:3, 1:1, 3:4, 9:16, +1 |
| generate_audioopt | BOOLEAN | true | — |
| return_last_frameopt | BOOLEAN | false | — |
| seedopt | INT | -1-1–2147483647 | — |
| api_configopt | ZHENZHEN_SEEDANCE2_CONFIG | — | |
| first_imageopt | IMAGE | — | |
| last_imageopt | IMAGE | — | |
| image1opt | IMAGE | — | |
| image2opt | IMAGE | — | |
| image3opt | IMAGE | — | |
| image4opt | IMAGE | — | |
| image5opt | IMAGE | — | |
| image6opt | IMAGE | — | |
| image7opt | IMAGE | — | |
| image8opt | IMAGE | — | |
| image9opt | IMAGE | — | |
| video1opt | VIDEO | — | |
| video2opt | VIDEO | — | |
| video3opt | VIDEO | — | |
| audio1opt | AUDIO | — | |
| audio2opt | AUDIO | — | |
| audio3opt | AUDIO | — | |
| skip_erroropt | BOOLEAN | false | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| task | COMFLY_VIDEO_FUTURE | — |