ToAPIsToAPIs
seedance-2-5

Seedance 2.5 flash Video Generation

Generate videos with the Seedance 2.5 flash model

POST
/v1/videos/generations

Note for users in mainland China: Please use https://toapis.cn as the API endpoint (Base URL). Replace https://toapis.com with https://toapis.cn in the examples in this document.

Seedance 2.5 flash is the faster, lower-cost variant of Seedance 2.5. It uses the public model name seedance-2-5-flash.

  • Supports text-to-video, first-frame / first-last-frame image-to-video, editing, video extension, and multimodal reference workflows.
  • Supports image, video, and audio references, including audio-only references.
  • The task is asynchronous. The submit API returns a generation.task; fetch the finished video from the task status API.

Compared with Seedance 2.5, flash generates about twice as fast at a lower price, outputs only 480p or 720p, accepts up to 20 reference items (10 images, 5 videos, 5 audio clips), and does not support draft mode.

Authorizations

Authorizationstringheaderrequired

Use Bearer Token authentication for every request.

Get an API key from the API Key Management Page.

Authorization: Bearer YOUR_API_KEY

Body

modelstringbodyrequireddefault: seedance-2-5-flash

Public ToAPIs model name. Use seedance-2-5-flash.

promptstringbody

Describe the scene, camera movement, subject action, style, and audio mood. When referring to media, use labels such as “image 1”, “video 1”, or “audio 1”.

video_operationstringbodydefault: generate

Video operation: generate, edit, or extend. Editing and extension require at least one video_with_roles; editing uses duration=-1, while extension accepts 4–30 or -1. All three operations use this endpoint.

client_business_idstringbody

Your order number or internal job ID. It is stored with the task and can be used for later lookup with GET /v1/videos/generations/{client_business_id}.

durationintegerbody

Output duration in seconds. The official range is 4–30; use -1 to let the model choose the duration. Editing accepts only -1, and the output length matches the input video. Extension accepts 4–30 or -1, measured as the total output duration after extension.

aspect_ratiostringbody

Output aspect ratio. Allowed values: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16, and adaptive.

Use adaptive for editing, video extension, and first/last-frame tasks, or when the input media should determine the ratio.

image_urlsstring[]body

Compatibility image URL array. For new integrations, prefer image_with_roles so each image has an explicit role. Do not send both fields in one request.

image_with_rolesarraybody

Images with explicit roles. Each item contains a required url and role.

Roles: first_frame, last_frame, or reference_image.

Up to 10 images may be supplied. First/last-frame mode and reference-image mode are mutually exclusive.

video_with_rolesarraybody

Reference videos. Each item contains a required url and the role reference_video. Up to 5 videos may be supplied.

audio_with_rolesarraybody

Reference audio. Each item contains a required url and the role reference_audio. Up to 5 audio clips may be supplied. Audio-only reference input is supported.

resolutionstringbody

Output resolution: 480p or 720p. Seedance 2.5 flash does not support 1080p or 4k.

output_formatstringbodydefault: mp4

Supported formats are mp4 and mov.

generate_audiobooleanbodydefault: true

Whether to generate synchronized audio.

return_last_framebooleanbodydefault: false

Whether to return the generated video's last-frame image in the completed task result.

toolsarraybody

Optional tools. The supported value is [{ "type": "web_search" }].

Web search is for text-to-video requests and cannot be combined with image, video, or audio inputs.

callback_urlstringbody

ToAPIs task-completion callback URL. See Task Webhooks.

trace_idstringbody

A caller-supplied tracing ID.

disable_nsfwbooleanbodydefault: true

NSFW content generation is disallowed by default. Set to false to allow it.

Reference input limits

The combined limit is 20 reference items: up to 10 images, 5 videos, and 5 audio clips. Use explicit roles to avoid ambiguous frame/reference interpretation.

Invalid task-type combinations can fail. Editing and extension tasks set the upstream task type explicitly, so a constraint mismatch returns an error immediately at submit time. For other task types the model infers the type from the media and the prompt, and a mismatch fails asynchronously with InvalidParameter.TaskTypeConstraint. For editing, extension, and first/last-frame tasks, set aspect_ratio to adaptive as required by the upstream API.

Asset review

When a request includes image, video, or audio inputs, some channels review the media automatically before submitting generation. No extra parameter is needed; pass media URLs in the existing media fields.

  • An accepted request returns a video task immediately. When review is required, the task stays queued and generation is submitted only after all media in the request pass. First-time review can take several minutes; reusing the same URL for the same user and channel reuses the review result.
  • Each media preparation round lasts up to 20 minutes by default. Rejection or preparation timeout ends the task as failed; see error.message for the reason. Failed review incurs no video generation fee.
  • Media must be public HTTP(S) URLs; Base64 and local paths are not accepted. Keep images within 30 MB and video or audio within 100 MiB. Each model's existing format limits still apply.
  • Channels that review media validate it at submission: a URL with embedded credentials or longer than 2048 bytes, the same URL used as different media types, or more than 64 media references in one request returns HTTP 400 with error.code invalid_private_assets.
  • Media is identified by URL. Use a new URL after the file content changes.
  • Legacy asset:// references are retired and return HTTP 400.

Fetch completed results from Video task status. When original URL delivery is enabled, prefer result.data[].url and use the optional fallback_url only if the primary URL cannot be accessed. Last frames use last_frame_url and fallback_last_frame_url. Fallback fields are returned only when copies are available.

Response

idstring
Task ID used to query status.
client_business_idstring
Returned when client_business_id was supplied.
objectstring
Fixed to generation.task.
modelstring
The requested model name.
statusstring
queued, in_progress, completed, or failed.
progressinteger
Progress percentage from 0 to 100.
created_atinteger
Creation timestamp.

requestExample

curl --request POST \
  --url https://toapis.com/v1/videos/generations \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "seedance-2-5-flash",
    "prompt": "A cinematic product shot, slow camera push-in, natural room tone.",
    "duration": 10,
    "aspect_ratio": "16:9",
    "resolution": "720p",
    "generate_audio": true
  }'
{
  "model": "seedance-2-5-flash",
  "prompt": "Use image 1 for the character, video 1 for motion, and audio 1 for rhythm.",
  "duration": 12,
  "aspect_ratio": "16:9",
  "image_with_roles": [{"url": "https://example.com/ref.png", "role": "reference_image"}],
  "video_with_roles": [{"url": "https://example.com/motion.mp4", "role": "reference_video"}],
  "audio_with_roles": [{"url": "https://example.com/rhythm.mp3", "role": "reference_audio"}]
}
{
  "model": "seedance-2-5-flash",
  "prompt": "Extend video 1 forward, keep the camera pushing in and reveal the city outside the window.",
  "video_operation": "extend",
  "duration": 11,
  "aspect_ratio": "adaptive",
  "output_format": "mov",
  "video_with_roles": [{"url": "https://example.com/source.mov", "role": "reference_video"}]
}

responseExample

{
  "id": "tsk_vid_xxx",
  "object": "generation.task",
  "model": "seedance-2-5-flash",
  "status": "in_progress",
  "progress": 10,
  "created_at": 1781577600
}