Seedance 2.5 Video Generation

Describe a scene in everyday words to make a Seedance 2.5 video. You can also share a photo, video, or sound link if your AI assistant can open it, and explain how it should guide the result.

Your AI

Use with your AI

For tasks with several steps, continue in the AI you already use.

CodexClaude CodeCursorWindsurfCline

View instructions

Use the Seedance 2.5 video tool available through my Toolmu setup to make one video that matches my request. I may describe a scene in words or share a photo, video, or sound link. Use a link only if you can open it; if you cannot, ask me for another link or offer a text-only request. If I say a photo should be the opening picture, start with it. If it should guide the subject or style, use it as a visual guide. If I request an exact ending picture, ask me for an opening picture too. When I ask to edit a clip, keep the same vertical or horizontal direction and length as the original; when I ask to continue it, keep that direction and describe only what happens next. Translate plain requests such as [vertical], [square], or [about 5 seconds] into the available choices. Ask me if a requested length or screen direction cannot be used. Follow my plain-language request for speech, music, or sound effects; if I ask for silence, leave out generated sound. Before anything that may cost money, show me the current price and wait for my approval. Follow the request until it completes or fails, then report what actually happened. Do not say a video is ready or give viewing or download steps unless the completed task actually provides them.

Task:
Make a short video of [a red fox running through autumn leaves]. Make it [vertical], last about [5 seconds], show [gentle camera movement], and include [no spoken dialogue].

Paste into your connected AI's conversation.

Connection guide

Up to 120 credits / call

The final charge is based on confirmed usage. Unused reserved credits return to your available balance.

ByteDance

Seedance 2.5 Video Generation

Describe a scene in everyday words to make a Seedance 2.5 video. You can also share a photo, video, or sound link if your AI assistant can open it, and explain how it should guide the result.

You receive

After successful completion, use the generated clip as a product-demo or visual-concept draft. You can request MP4 or MOV; viewing or downloading depends on what the completed task returns.

Start with your task

Make a short video of [a red fox running through autumn leaves]. Make it [vertical], last about [5 seconds], show [gentle camera movement], and include [no spoken dialogue].

What to know

For editing, request automatic length. Using opening or closing frames, editing or extending a clip requires the source-matching ratio setting. References do not guarantee exact reproduction, and submitting a task does not mean its video is finished.

Inputs & outputs

Inspect the current inputs, availability, and maximum charge before starting.

Input schema

Input schema
Field / Path	Type	Requirement	Meaning
Video request`body`	`object`	Required	Contains the content items and optional video, sound and format choices. Required fields `content`
Aspect ratio`body.ratio`	`string`	Optional	Choose a fixed ratio for text or reference-guided generation. Use adaptive for opening/closing frames, editing or extension so the request follows the source ratio. Default `adaptive` Options `16:9, 4:3, 1:1, 3:4, 9:16, 21:9, adaptive`
Scene and reference items`body.content`	`array`	Required	Provide at least one text, image, video or audio item. Describe the intended task and what each reference should contribute; text is not required for every request. Minimum items `1`
Content item`body.content[]`	`object`	—	Each item is one text prompt or one image, video or audio reference. Supply the matching text or nested media URL.
Scene description`body.content[].text`oneOf · 1	`string`	—	For a text item, write 1–6,000 characters. Use @Image1, @Video1 or @Audio1 to identify references in their per-type order and explain what to retain or change. Minimum length `1` Maximum length `6000`
Item type`body.content[].type`oneOf · 1	`string`	—	Choose text, image_url, video_url or audio_url for the item, then provide that type's text or media URL. Fixed value `text`
Reference role`body.content[].role`oneOf · 2	`string`	—	Use first_frame or last_frame for opening/closing images, with a first frame when supplying a last frame. Use reference_image for appearance/style, reference_video for video guidance, or reference_audio for sound; audio may be the only reference. Options `first_frame, last_frame, reference_image`
Item type`body.content[].type`oneOf · 2	`string`	—	Choose text, image_url, video_url or audio_url for the item, then provide that type's text or media URL. Fixed value `image_url`
Image reference`body.content[].image_url`oneOf · 2	`object`	—	Contains the image URL. Choose whether the image guides appearance/style or acts as an opening or closing frame. Required fields `url`
Image URL`body.content[].image_url.url`oneOf · 2	`string`	—	Provide a nonempty link to an image the provider can access. Appearance/style requests can use up to 30 reference images. Minimum length `1`
Reference role`body.content[].role`oneOf · 3	`string`	—	Use first_frame or last_frame for opening/closing images, with a first frame when supplying a last frame. Use reference_image for appearance/style, reference_video for video guidance, or reference_audio for sound; audio may be the only reference. Fixed value `reference_video`
Item type`body.content[].type`oneOf · 3	`string`	—	Choose text, image_url, video_url or audio_url for the item, then provide that type's text or media URL. Fixed value `video_url`
Video reference`body.content[].video_url`oneOf · 3	`object`	—	Contains a clip URL for motion guidance, editing or extension. Explain which parts of the clip should inform the result. Required fields `url`
Video URL`body.content[].video_url.url`oneOf · 3	`string`	—	Provide an accessible clip link. Use up to 10 clips, each 2–30 seconds, with a combined length no greater than 30 seconds. Minimum length `1`
Reference role`body.content[].role`oneOf · 4	`string`	—	Use first_frame or last_frame for opening/closing images, with a first frame when supplying a last frame. Use reference_image for appearance/style, reference_video for video guidance, or reference_audio for sound; audio may be the only reference. Fixed value `reference_audio`
Item type`body.content[].type`oneOf · 4	`string`	—	Choose text, image_url, video_url or audio_url for the item, then provide that type's text or media URL. Fixed value `audio_url`
Audio reference`body.content[].audio_url`oneOf · 4	`object`	—	Contains sound or timbre guidance. It may be the only reference and is separate from requesting newly generated audio. Required fields `url`
Audio URL`body.content[].audio_url.url`oneOf · 4	`string`	—	Provide an accessible audio link. Use up to 10 clips, each 2–30 seconds, with a combined length no greater than 30 seconds. Minimum length `1`
Requested length`body.duration`	`integer \| string`	Optional	Choose an integer from 4 to 30 seconds or auto. For video editing, use auto to request a length that follows the source clip. Default `5`
Queue priority`body.priority`	`integer`	Optional	Choose an integer from 0 to 9. Higher values request greater queue priority; they do not specify a completion time. Minimum `0` Maximum `9`
AI watermark`body.watermark`	`boolean`	Optional	Set true to request an AI Generated watermark, or false to request its omission. Default `false`
Requested resolution`body.resolution`	`string`	Optional	Choose 480p or 720p and check the current charge for that choice before submitting. Default `720p` Options `480p, 720p`
Requested format`body.output_format`	`string`	Optional	Choose mp4 for common playback workflows, or mov for editing and colour work. MOV may require a separate player or editor; actual access depends on the completed result. Default `mp4` Options `mp4, mov`
Generated sound`body.generate_audio`	`boolean`	Optional	Set true to request synchronized voice, effects or music; false requests no generated sound. A supplied audio reference is a separate choice. Default `true`

View raw JSON

{
  "type": "object",
  "required": [
    "body"
  ],
  "properties": {
    "body": {
      "type": "object",
      "$schema": "https://json-schema.org/draft/2020-12/schema",
      "required": [
        "content"
      ],
      "~standard": {
        "vendor": "zod",
        "version": 1,
        "jsonSchema": {}
      },
      "properties": {
        "ratio": {
          "enum": [
            "16:9",
            "4:3",
            "1:1",
            "3:4",
            "9:16",
            "21:9",
            "adaptive"
          ],
          "type": "string",
          "default": "adaptive",
          "description": "\"adaptive\" (the default) matches your inputs and is ALWAYS safe. A specific ratio is accepted ONLY for text-to-video and reference-to-video. First-frame/last-frame, video editing, and video extension all inherit the aspect ratio of their source and REQUIRE \"adaptive\" - anything else is rejected (up front for first-frame/last-frame, and by the model after queueing for editing and extension)."
        },
        "content": {
          "type": "array",
          "items": {
            "oneOf": [
              {
                "type": "object",
                "required": [
                  "type",
                  "text"
                ],
                "properties": {
                  "text": {
                    "type": "string",
                    "maxLength": 6000,
                    "minLength": 1,
                    "description": "Text prompt describing the video (recommended <1000 words). Refer to your other content items by ordinal, numbered per type in array order: @Image1, @Image2, @Video1, @Audio1 - and say what each one supplies (appearance, motion, timbre) and what it should not. Mark sounds with () music, <> sound effects, {} dialogue, 【】 subtitles; prefix non-English dialogue with its language."
                  },
                  "type": {
                    "type": "string",
                    "const": "text"
                  }
                },
                "additionalProperties": false
              },
              {
                "type": "object",
                "required": [
                  "type",
                  "image_url"
                ],
                "properties": {
                  "role": {
                    "enum": [
                      "first_frame",
                      "last_frame",
                      "reference_image"
                    ],
                    "type": "string",
                    "description": "first_frame (image-to-video, default) pins the exact opening frame; last_frame (only with a first_frame) pins the closing one; reference_image (up to 30) is soft guidance for subject or style, not a frame."
                  },
                  "type": {
                    "type": "string",
                    "const": "image_url"
                  },
                  "image_url": {
                    "type": "object",
                    "required": [
                      "url"
                    ],
                    "properties": {
                      "url": {
                        "type": "string",
                        "minLength": 1
                      }
                    },
                    "additionalProperties": false
                  }
                },
                "additionalProperties": false
              },
              {
                "type": "object",
                "required": [
                  "type",
                  "video_url"
                ],
                "properties": {
                  "role": {
                    "type": "string",
                    "const": "reference_video",
                    "description": "Reference video (up to 10, each 2-30s, total <=30s) for multimodal generation, video editing, or extension."
                  },
                  "type": {
                    "type": "string",
                    "const": "video_url"
                  },
                  "video_url": {
                    "type": "object",
                    "required": [
                      "url"
                    ],
                    "properties": {
                      "url": {
                        "type": "string",
                        "minLength": 1
                      }
                    },
                    "additionalProperties": false
                  }
                },
                "additionalProperties": false
              },
              {
                "type": "object",
                "required": [
                  "type",
                  "audio_url"
                ],
                "properties": {
                  "role": {
                    "type": "string",
                    "const": "reference_audio",
                    "description": "Reference audio (up to 10, each 2-30s, total <=30s). May be the only reference."
                  },
                  "type": {
                    "type": "string",
                    "const": "audio_url"
                  },
                  "audio_url": {
                    "type": "object",
                    "required": [
                      "url"
                    ],
                    "properties": {
                      "url": {
                        "type": "string",
                        "minLength": 1
                      }
                    },
                    "additionalProperties": false
                  }
                },
                "additionalProperties": false
              }
            ]
          },
          "minItems": 1,
          "description": "Content items: text (text-to-video), image_url (image-to-video / first+last frame / reference images), video_url and audio_url (multimodal reference). The model infers WHICH task you want from these roles plus your prompt wording - see the endpoint notes."
        },
        "duration": {
          "anyOf": [
            {
              "type": "integer",
              "maximum": 30,
              "minimum": 4
            },
            {
              "type": "string",
              "const": "auto"
            }
          ],
          "default": 5,
          "description": "Duration in seconds (4-30), or \"auto\" to let the model choose the length from your prompt and inputs. \"auto\" is REQUIRED for video editing (the output matches the source clip and no explicit duration is accepted). Cost note: \"auto\" holds the price of a 30s video up front - the unused portion is released when the run settles on actual usage."
        },
        "priority": {
          "type": "integer",
          "maximum": 9,
          "minimum": 0,
          "description": "Queue priority 0-9 (higher = front of the queue). Default 0."
        },
        "watermark": {
          "type": "boolean",
          "default": false,
          "description": "Add an \"AI Generated\" watermark. Default false."
        },
        "resolution": {
          "enum": [
            "480p",
            "720p"
          ],
          "type": "string",
          "default": "720p",
          "description": "Output resolution (also the price selector)."
        },
        "output_format": {
          "enum": [
            "mp4",
            "mov"
          ],
          "type": "string",
          "default": "mp4",
          "description": "\"mp4\" (default) plays everywhere - web, mobile, players, social platforms. \"mov\" uses H.264 + yuv444p + PCM for higher colour fidelity in grading, keying, and compositing (recommended for video editing and extension), but most browsers cannot play it - use VLC, mpv, ffplay, or IINA."
        },
        "generate_audio": {
          "type": "boolean",
          "default": true,
          "description": "Generate synchronized audio (voice, sound effects, music). Default true. No price impact."
        }
      },
      "additionalProperties": false
    }
  },
  "additionalProperties": false
}

Result schema

Output fields have not been documented yet.

Additional notes

Give each reference a job

For an appearance reference, explain which subject details or style to retain and what may change. Number images, videos and audio separately in the order supplied, and refer to those numbers in the scene description. An opening or closing image serves a different purpose from a general appearance reference.

Plan edits and playback

When editing a clip, describe what should change and what should stay, and choose automatic duration. For an extension, describe the continuation you want. Opening/closing-frame requests, editing and extension use the adaptive ratio setting to follow the source material.

If you choose MOV for colour work, plan for a player or editor that accepts H.264 video, yuv444p colour and PCM audio. MP4 is the common playback request. Check what the completed task returns before relying on a particular viewing or download step.

Seedance 2.5 Video Generation

Seedance 2.5 Video Generation

You receive

Start with your task

What to know

Inputs & outputs

Input schema

Result schema

Give each reference a job

Plan edits and playback

Use your AI assistant

Describe the scene

Guide, edit, or continue a clip