seedance-2-5

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Seedance 2.5

Seedance 2.5 extends the Seedance 2 family to 4–30 second 480p/720p clips and larger multimodal reference sets. It is hosted; there are no local model weights in OpenMontage.

Seedance 2.5 将Seedance 2系列扩展至4–30秒的480p/720p视频片段，并支持更大规模的多模态参考集。该模型为托管式，OpenMontage中无本地模型权重。

Choose a supported route

选择支持的调用渠道

Route	Tool call	Notes
fal.ai	`seedance_video` , `model_version: "2.5"`	T2V, I2V, and reference-to-video
Volcengine Ark	`seedance_ark` , `model: "2.5"`	First-party model ID `doubao-seedance-2-5-260628` ; custom token price required for cost estimates
Runway	`runway_video` , `model: "seedance2_5"`	T2V, I2V, V2V; 480p/720p
ComfyUI Partner Node	`comfyui_video` , `model_family: "seedance_2.5"`	Hosted and paid despite running in a ComfyUI graph

Do not invent a Replicate, HeyGen, or Higgsfield identifier when their current public API schema does not list Seedance 2.5.

渠道	工具调用	说明
fal.ai	`seedance_video` , `model_version: "2.5"`	支持文本转视频（T2V）、图片转视频（I2V）以及参考内容转视频
Volcengine Ark	`seedance_ark` , `model: "2.5"`	官方模型ID为 `doubao-seedance-2-5-260628` ；成本估算需自定义 token 价格
Runway	`runway_video` , `model: "seedance2_5"`	支持文本转视频（T2V）、图片转视频（I2V）、视频转视频（V2V）；分辨率为480p/720p
ComfyUI Partner Node	`comfyui_video` , `model_family: "seedance_2.5"`	虽在ComfyUI图中运行，但属于托管付费服务

若Replicate、HeyGen或Higgsfield的当前公开API架构未列出Seedance 2.5，请勿自行编造其调用标识。

Reference limits

参考内容限制

Up to 30 reference images.
Up to 10 reference videos.
Up to 10 reference audio clips.
Keep combined reference video/audio duration within the provider's documented ceiling; Runway caps it at 30 seconds.
For Runway video-to-video, the source video consumes one video slot.

Reference inputs are provider-specific. fal.ai uses

image_urls

video_urls

, and

audio_urls

internally. Runway uses

references

referenceVideos

, and

referenceAudio

. Ark uses typed content entries with roles. Always call the OpenMontage tool instead of constructing a provider payload manually.

最多支持30张参考图片
最多支持10个参考视频
最多支持10个参考音频片段
参考视频与音频的总时长需在服务商规定上限内；Runway的上限为30秒
对于Runway的视频转视频功能，源视频将占用一个视频配额

参考输入格式因服务商而异。fal.ai内部使用

image_urls

、

video_urls

和

audio_urls

；Runway使用

references

、

referenceVideos

和

referenceAudio

；Ark使用带角色的类型化内容条目。请始终调用OpenMontage工具，而非手动构建服务商请求 payload。

Prompting

提示词撰写

Lead with the shot structure, then subject, action, camera, lighting, and audio. For multi-shot work, give each beat an explicit time range and use quoted text for dialogue. Reference each supplied asset by a stable role in the prompt. Thirty seconds is a ceiling, not a target: use a shorter generation when the scene has only one meaningful action.

先描述镜头结构，再依次说明主体、动作、镜头、灯光和音频。对于多镜头作品，为每个节拍指定明确的时间范围，并使用引号标注对话。在提示词中通过固定角色引用每个提供的素材。30秒是上限而非目标：若场景仅包含一个核心动作，可生成更短的视频。

Cost and verification

成本与验证

All supported routes are paid. Confirm the exact provider/model before calling and review the result for identity continuity, cuts, lip sync, audio artifacts, and prompt adherence. ComfyUI Partner Nodes use prepaid Comfy credits and are not an offline fallback.

所有支持的渠道均为付费服务。调用前请确认具体的服务商/模型，并检查生成结果的角色一致性、剪辑衔接、唇形同步、音频瑕疵以及提示词匹配度。ComfyUI Partner Node使用预付费的Comfy credits，并非离线备用方案。