Seedance 2.0
ByteDance Seedance video generation. Defaults to Seedance 2.0 (set `quality`: 2.0 | 2.0-fast | 2.0-mini | 1.5-pro | lite | pro | pro-fast). Does text-to-video AND image-to-video with a first frame (`image_url`) plus optional last frame (`end_image_url`), reference images, and (on the 2.0 tier) a reference VIDEO via `video_url`, capped at 15s of footage (trim longer clips first, they are refused not truncated). Aliases: seedance_2, seedance_pro, seedance_i2v, seedance_2_i2v.
Try Seedance 2.0 in the playground
Seedance 2.0
Badge: New
Node type: video:bytedance_seedance_v1_lite_text_to_video
Category: video
Description
Seedance 2.0 is ByteDance's video generation model, and this is the main Seedance node in Wireflow. It defaults to Seedance 2.0; set Version (quality) to 2.0, 2.0-fast, 2.0-mini, 1.5-pro, lite, pro or pro-fast to switch. It does text to video and image to video with a first frame (image_url) plus an optional last frame (end_image_url), reference images, and on the 2.0 tier a reference video via video_url, capped at 15 seconds of footage (trim longer clips first; they are refused, not truncated). Aliases: seedance_2, seedance_pro, seedance_i2v, seedance_2_i2v. Its node type is video:bytedance_seedance_v1_lite_text_to_video. Call it from code by running a saved workflow that contains it with POST /api/v1/workflows/{id}/run, or let an agent build and run it over the Wireflow MCP server. This page prices it at about 1 credit per video token.
Pricing
- Cost: ~1 credit per video-token
Variant family
- Family:
bytedance_seedance - Is primary: yes
- All variants:
video:seedance_2_pro,video:seedance_2_0_t2v,video:seedance_2_0_i2v,video:seedance_2_0_r2v,video:seedance_2_0_fast_t2v,video:seedance_2_0_fast_i2v,video:seedance_2_0_mini_t2v,video:seedance_2_0_mini_i2v,video:seedance_2_0_mini_r2v,video:bytedance_seedance_v1_5_pro_i2v,video:bytedance_seedance_v1_5_pro_t2v,video:bytedance_seedance_v1_lite_image_to_video,video:bytedance_seedance_v1_lite_text_to_video
Canvas ports
These appear as port handles on the left side of the node.
| ID | Label | Details |
|---|---|---|
prompt |
Prompt | TEXT (required) |
image_url |
Start Frame | IMAGE |
end_image_url |
End Frame | IMAGE |
reference_image_urls |
Reference Images | IMAGE |
video_url |
Reference Video | VIDEO |
audio_urls |
Reference Audio | AUDIO — A guide for the voice, not the voice itself: the track never plays in the result. With Generate Audio on, Seedance reads the timbre, timing and words from this track and then generates its own audio, so the lips will not match it exactly. Turning Generate Audio off gives a silent video, it does not keep this track. On version 2.5, in an edit (a Reference Video wired and a prompt that changes it), wire the audio of that same clip here, turn Generate Audio off, and add this line to the prompt: "Match the framing of @Video1 in every frame: same face size and position, same distance from the camera. Never move closer or crop tighter." Without it, the audio can pull a wide shot into a close-up. The mouth then follows the timing and big shapes of this track, close but not frame perfect; check the take length against the clip and retime it in the Video Editor if it is off, then lay the audio back in. Keep Generate Audio off when the track is a copyrighted song, because with it on Seedance can refuse the generated audio for copyright. Read by versions 2.5 and 2.0, and by 2.0 Mini only beside Reference Images; the other versions ignore it. |
Sidebar config
These render as form fields in the right-side config panel when the node is selected.
| ID | Label | Details |
|---|---|---|
quality |
Version | TEXT · options: 2.0, 2.0-fast, 2.0-mini, 1.5-pro, lite, +2 |
aspect_ratio |
Aspect Ratio | TEXT · options: 21:9, 16:9, 4:3, 1:1, 3:4, +2 |
camera_fixed |
Camera Fixed | BOOLEAN |
duration |
Duration (seconds) | TEXT · options: auto, 4, 5, 6, 7, +8 |
resolution |
Resolution | TEXT · options: 480p, 720p, 1080p |
generate_audio |
Generate Audio | BOOLEAN |
contentFilter |
Content filter | BOOLEAN · default true. Applies to the Seedance 2.5 and 2.0 quality tiers. On by default. Seedance screens your prompt, your reference material and the finished clip, so a run can fail after it has been submitted when something gets flagged (copyright is the common one). Turning this off asks Seedance to skip that safety and IP moderation. It costs 10 percent more per run, and you are responsible for holding the rights to whatever you generate. |
Content filter, and what turning it off costs
On the 2.5 and 2.0 quality tiers Seedance moderates in three separate places: your prompt, the reference material you register, and the FINISHED CLIP. The last one is the surprising one. A run can pass the first two checks, generate, and then fail after submission with a message like "the request failed because the output video may be related to copyright restrictions".
Content filter is on by default, which is the behaviour every existing board
already has. Switch it off and Wireflow sends content_filter: false, asking
Seedance to skip that safety and IP moderation.
- It costs 10 percent more. The quote on the card, the credit gate, the charge and the final settlement all go up by 10 percent, so the number you see is the number you pay.
- The rights are yours to hold. Skipping the filter licenses nothing. You are responsible for the material you feed in and for whatever comes out.
On the 1.5-pro, lite and pro tiers there is no such setting, so the field does nothing there and nothing extra is charged.
Outputs
| ID | Label | Type |
|---|---|---|
video |
Generated Video | VIDEO |
Reference Video is a 2.0-tier input, capped at 15s
Wire a clip into the Reference Video port to condition the generation on real footage (motion, framing, style). It only applies on the 2.0 version (the default). Other versions route to models that do not take a reference video, so a clip wired there is ignored.
Seedance caps reference footage at 15 seconds. A longer clip is refused before you are charged, with the measured length in the message, so trim it upstream first. It is never silently truncated. The duration is read from the file itself, not from any number saved on the upstream node, so the cap cannot be dodged by editing config.
Billing note: a run with a reference video reserves credits for a full 15s of footage up front, then settles DOWN to what the model actually used. The quote you see is the upfront reservation, not the final cost.
One node, every Seedance version
This is the single Seedance node. Version picks the tier (2.0, 2.0 Fast, 2.0 Mini, 1.5 Pro, v1 Lite/Pro), and the same ports drive all of them: a Prompt for text-to-video, a Start Frame (plus optional End Frame) for image-to-video, Reference Images, and on 2.0 the Reference Video above. Leave the frames unwired for pure text-to-video.
Auto-generated from the Wireflow node registry.