> ## Documentation Index > Fetch the complete documentation index at: https://apidoc.cometapi.com/llms.txt > Use this file to discover all available pages before exploring further. # Tạo một video Seedance > Tạo một tác vụ video Seedance trên CometAPI. Seedance 2.0 hỗ trợ các giá trị WxH chính xác lên đến 4K để tạo video từ văn bản và ảnh tham chiếu. Chọn một giá trị `WxH` được ghi trong tài liệu từ bảng bên dưới cho trường `size`. ## Sử dụng ảnh tham chiếu Chỉ các mô hình Seedance 2.0 mới chấp nhận `input_reference`. Thêm một tệp `input_reference` để tạo nội dung có hướng dẫn bằng ảnh. Đối với các Prompt nhiều ảnh tham chiếu nâng cao, hãy lặp lại cùng một trường multipart theo thứ tự tải lên. Seedance 2.0 nhạy với cách Prompt liên kết từng ảnh tham chiếu với cảnh. Hãy tham chiếu đến các tệp đã tải lên theo thứ tự, chẳng hạn `[Image 1]`, `[Image 2]` và `[Image 3]`, rồi gán cho mỗi ảnh một vai trò rõ ràng: * Dùng `[Image 1]` cho chủ thể chính, nhân vật hoặc sản phẩm. * Dùng `[Image 2]` cho chủ thể phụ, đạo cụ hoặc bạn đồng hành. * Dùng `[Image 3]` cho nền, bối cảnh, ánh sáng hoặc phong cách. Mô tả những gì cần vẫn dễ nhận biết, những gì có thể thay đổi, hành động, chuyển động máy quay, phong cách hình ảnh và cảnh. Một Prompt mơ hồ có thể khiến video được tạo trông như thể ảnh tham chiếu không được sử dụng, ngay cả khi tệp tải lên đã được chấp nhận. Cấu trúc Prompt này đáng tin cậy hơn việc chỉ liệt kê ảnh: ```text theme={null} Use the uploaded images in order: keep the boy wearing glasses and a blue T-shirt from [Image 1], add the corgi puppy from [Image 2], and use the lawn from [Image 3] as the setting. Create a 4-second 3D cartoon video with the boy and puppy sitting together, a slow camera push-in, soft daylight, and no extra characters. ``` ## Thời lượng theo mô hình Nếu bỏ qua `seconds`, CometAPI sẽ yêu cầu một clip dài 5 giây. Gửi `seconds` dưới dạng chuỗi khi bạn cần thời lượng cụ thể. | Dòng mô hình | Được hỗ trợ `seconds` | Mặc định | Hành vi ở giới hạn | | --------------------------------- | --------------------- | -------- | ------------------------------------- | | Seedance 2.0 và Seedance 2.0 Fast | `4`-`15` | `5` | Sử dụng số giây nguyên trong phạm vi. | | Seedance 1.5 Pro | `4`-`12` | `5` | Sử dụng số giây nguyên trong phạm vi. | | Seedance 1.0 Pro | `2`-`10` | `5` | Sử dụng số giây nguyên trong phạm vi. | ## Hỗ trợ kích thước theo model Bảng dưới đây tuân theo ánh xạ độ phân giải Seedance chính thức và tập hợp các giá trị kích thước đã công bố tại một nơi. Đối với `1080p`, cột ngoài cùng bên phải liệt kê các giá trị `Seedance 1.5 Pro` và `Seedance 2.0` đã được ghi nhận. `Seedance 2.0 Fast` không được liệt kê trong bảng `1080p` chính thức đó. Các mục `4K` chỉ áp dụng cho `doubao-seedance-2-0`. Bạn có thể dùng chúng cho yêu cầu chuyển văn bản thành video hoặc chuyển hình ảnh thành video có bao gồm `input_reference`. Chúng không áp dụng cho `doubao-seedance-2-0-fast`. Hãy truyền chính xác giá trị `WxH` hiển thị trong bảng. | Lớp độ phân giải | Tỷ lệ khung hình | Giá trị pixel cho dòng Seedance 1.0 | Giá trị pixel cho Seedance 1.5 Pro / Seedance 2.0 / Seedance 2.0 Fast | | ---------------- | ---------------- | ----------------------------------- | --------------------------------------------------------------------- | | `480p` | `16:9` | `864x480` | `864x496` | | | `4:3` | `736x544` | `752x560` | | | `1:1` | `640x640` | `640x640` | | | `3:4` | `544x736` | `560x752` | | | `9:16` | `480x864` | `496x864` | | | `21:9` | `960x416` | `992x432` | | `720p` | `16:9` | `1248x704` | `1280x720` | | | `4:3` | `1120x832` | `1112x834` | | | `1:1` | `960x960` | `960x960` | | | `3:4` | `832x1120` | `834x1112` | | | `9:16` | `704x1248` | `720x1280` | | | `21:9` | `1504x640` | `1470x630` | | `1080p` | `16:9` | `1920x1088` | `1920x1080` | | | `4:3` | `1664x1248` | `1664x1248` | | | `1:1` | `1440x1440` | `1440x1440` | | | `3:4` | `1248x1664` | `1248x1664` | | | `9:16` | `1088x1920` | `1080x1920` | | | `21:9` | `2176x928` | `2206x946` | | `4K` | `16:9` | — | `3840x2160` (Chỉ dành cho Seedance 2.0) | | | `4:3` | — | `3326x2494` (Chỉ dành cho Seedance 2.0) | | | `1:1` | — | `2880x2880` (Chỉ dành cho Seedance 2.0) | | | `3:4` | — | `2494x3326` (Chỉ dành cho Seedance 2.0) | | | `9:16` | — | `2160x3840` (Chỉ dành cho Seedance 2.0) | | | `21:9` | — | `4398x1886` (Chỉ dành cho Seedance 2.0) | Chọn giá trị `WxH` đã được ghi nhận cho model đích. Hỗ trợ `WxH` chính xác vẫn phụ thuộc vào model, vì vậy hãy kiểm tra media đã hoàn tất trước khi dựa vào một kích thước chưa được ghi nhận. ## OpenAPI ````yaml api/openapi/video/seedance/post-seedance-create.openapi.json POST /v1/videos openapi: 3.1.0 info: title: Seedance Video Generation API version: 1.0.0 description: >- Create an asynchronous ByteDance Seedance video task through CometAPI. Seedance 2.0 models also accept one or more input_reference image files; for multiple files, repeat the same multipart field and bind each image in the prompt as [Image 1], [Image 2], and so on. Set size to an exact WxH value such as 1280x720. The standard doubao-seedance-2-0 model accepts these exact 4K sizes for text-to-video and reference-image generation: 3840x2160, 3326x2494, 2880x2880, 2494x3326, 2160x3840, and 4398x1886. These 4K sizes do not apply to doubao-seedance-2-0-fast. Save the returned id and poll GET /v1/videos/{id} until the status reaches completed. servers: - url: https://api.cometapi.com security: - bearerAuth: [] paths: /v1/videos: post: summary: Create a Seedance video task description: >- Submit a text-to-video or image-to-video job for any Seedance model family. The request body is multipart/form-data so that optional input_reference images can be uploaded in the same call. Set size to an exact WxH value. The standard doubao-seedance-2-0 model accepts these six exact 4K sizes for text-to-video and reference-image generation: 3840x2160, 3326x2494, 2880x2880, 2494x3326, 2160x3840, and 4398x1886. These 4K sizes do not apply to doubao-seedance-2-0-fast. Image-to-video is supported only by the Seedance 2.0 models (doubao-seedance-2-0, doubao-seedance-2-0-fast); the 1.0 Pro and 1.5 Pro models accept text prompts only. For multiple reference images, repeat input_reference in upload order and describe each image's role in the prompt with labels such as [Image 1], [Image 2], and [Image 3]. The endpoint returns immediately with a task id; use GET /v1/videos/{id} to read the final result. operationId: seedance_create_video requestBody: required: true content: multipart/form-data: schema: type: object required: - prompt - model properties: prompt: type: string description: >- Text prompt that describes the video. Required. When using reference images, describe what each uploaded image should control, such as the subject from [Image 1], the secondary object from [Image 2], and the background or style from [Image 3]. Also state the action, camera motion, visual style, and scene. example: >- A slow cinematic camera push across a coastal landscape at sunrise. model: type: string description: >- Seedance model id. Only the two 2.0 models accept input_reference. enum: - doubao-seedance-2-0 - doubao-seedance-2-0-fast - doubao-seedance-1-5-pro - doubao-seedance-1-0-pro example: doubao-seedance-2-0 seconds: type: integer description: >- Video duration in seconds. The accepted range depends on the model: doubao-seedance-2-0 and doubao-seedance-2-0-fast accept 4 to 15, doubao-seedance-1-5-pro accepts 4 to 12, and doubao-seedance-1-0-pro accepts 2 to 10. The default is 5 for every model. minimum: 2 maximum: 15 default: 5 example: 5 size: type: string description: >- Output size as an exact WxH value, such as 1280x720. For doubao-seedance-2-0 4K output, use 3840x2160, 3326x2494, 2880x2880, 2494x3326, 2160x3840, or 4398x1886. These 4K values work for text-to-video and requests with input_reference, but they do not apply to doubao-seedance-2-0-fast. Exact WxH support remains model-dependent, and an undocumented value can normalize to another size or fail. pattern: ^[1-9]\d{2,3}x[1-9]\d{2,3}$ examples: - 1280x720 - 1920x1080 - 3840x2160 - 3326x2494 - 2880x2880 - 2494x3326 - 2160x3840 - 4398x1886 - 1112x834 - 960x960 - 834x1112 - 720x1280 - 1080x1920 - 1470x630 example: 1280x720 input_reference: type: string format: binary description: >- Optional reference image uploaded as a multipart file. Repeat this field to send multiple reference images; the order of repeated fields is the order you should reference in the prompt as [Image 1], [Image 2], and so on. Use JPEG or PNG files for best compatibility. Only doubao-seedance-2-0 and doubao-seedance-2-0-fast accept this field; sending it with a 1.0 Pro or 1.5 Pro model returns HTTP 400. encoding: input_reference: contentType: image/jpeg, image/png, image/webp examples: seedance_2_0_4k_text_to_video: summary: Seedance 2.0 4K text-to-video value: prompt: >- A slow cinematic camera push across a coastal landscape at sunrise. model: doubao-seedance-2-0 seconds: 4 size: 3840x2160 responses: '200': description: Task created. Save the returned id and poll GET /v1/videos/{id}. content: application/json: schema: $ref: '#/components/schemas/CreateVideoResponse' example: id: task_abc123 task_id: task_abc123 object: video model: doubao-seedance-2-0 status: queued progress: 0 created_at: 1776681149 '400': description: >- The request is missing a required field or contains an out-of-range value. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' examples: missing_prompt: summary: prompt field missing value: code: invalid_request message: prompt is required missing_model: summary: model field missing value: error: code: '' message: model name is required type: comet_api_error invalid_duration: summary: seconds outside the per-model range value: code: InvalidParameter message: >- the parameter duration specified in the request is not valid i2v_not_supported: summary: input_reference sent with a 1.0 Pro or 1.5 Pro model value: code: InvalidParameter message: >- The parameter task_type specified in the request is not valid: the specified task_type r2v does not support model seedance-1-5-pro '401': description: The API key is missing or invalid. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' examples: invalid_token: summary: Bearer token rejected value: error: code: '' message: invalid token type: comet_api_error x-codeSamples: - lang: Shell label: Text-to-video 4K (Seedance 2.0, exact 3840x2160) source: |- curl https://api.cometapi.com/v1/videos \ -H "Authorization: Bearer $COMETAPI_KEY" \ -F 'prompt="A slow cinematic camera push across a coastal landscape at sunrise"' \ -F 'model="doubao-seedance-2-0"' \ -F 'seconds="4"' \ -F 'size="3840x2160"' - lang: Python label: Text-to-video 4K (Seedance 2.0, exact 3840x2160) source: | import os import requests response = requests.post( "https://api.cometapi.com/v1/videos", headers={"Authorization": "Bearer " + os.environ["COMETAPI_KEY"]}, data={ "prompt": "A slow cinematic camera push across a coastal landscape at sunrise", "model": "doubao-seedance-2-0", "seconds": "4", "size": "3840x2160", }, timeout=30, ) response.raise_for_status() print(response.json()) - lang: JavaScript label: Text-to-video 4K (Seedance 2.0, exact 3840x2160) source: > const form = new FormData(); form.append("prompt", "A slow cinematic camera push across a coastal landscape at sunrise"); form.append("model", "doubao-seedance-2-0"); form.append("seconds", "4"); form.append("size", "3840x2160"); const response = await fetch("https://api.cometapi.com/v1/videos", { method: "POST", headers: { Authorization: `Bearer ${process.env.COMETAPI_KEY}` }, body: form, }); const result = await response.json(); console.log(result); - lang: Shell label: Single reference image (Seedance 2.0) source: |- curl https://api.cometapi.com/v1/videos \ -H "Authorization: Bearer $COMETAPI_KEY" \ -F 'prompt="Use [Image 1] as the main visual reference. Preserve the subject identity, clothing colors, and overall composition from [Image 1]. Create a 4-second video with a gentle camera push-in, subtle natural motion, warm daylight, and a realistic cinematic style."' \ -F 'model="doubao-seedance-2-0"' \ -F 'seconds="4"' \ -F 'size="720x1280"' \ -F 'input_reference=@"/path/to/reference.png"' - lang: Python label: Single reference image (Seedance 2.0) source: > import os import requests prompt = "Use [Image 1] as the main visual reference. Preserve the subject identity, clothing colors, and overall composition from [Image 1]. Create a 4-second video with a gentle camera push-in, subtle natural motion, warm daylight, and a realistic cinematic style." with open("/path/to/reference.png", "rb") as image: response = requests.post( "https://api.cometapi.com/v1/videos", headers={"Authorization": "Bearer " + os.environ["COMETAPI_KEY"]}, data={ "prompt": prompt, "model": "doubao-seedance-2-0", "seconds": "4", "size": "720x1280", }, files=[("input_reference", ("reference.png", image, "image/png"))], timeout=30, ) response.raise_for_status() print(response.json()) - lang: JavaScript label: Single reference image (Seedance 2.0) source: > import { readFile } from "node:fs/promises"; const prompt = "Use [Image 1] as the main visual reference. Preserve the subject identity, clothing colors, and overall composition from [Image 1]. Create a 4-second video with a gentle camera push-in, subtle natural motion, warm daylight, and a realistic cinematic style."; const image = await readFile("/path/to/reference.png"); const form = new FormData(); form.append("prompt", prompt); form.append("model", "doubao-seedance-2-0"); form.append("seconds", "4"); form.append("size", "720x1280"); form.append("input_reference", new Blob([image], { type: "image/png" }), "reference.png"); const response = await fetch("https://api.cometapi.com/v1/videos", { method: "POST", headers: { Authorization: `Bearer ${process.env.COMETAPI_KEY}` }, body: form, }); const result = await response.json(); console.log(result); - lang: Shell label: Multiple reference images (Seedance 2.0) source: |- curl https://api.cometapi.com/v1/videos \ -H "Authorization: Bearer $COMETAPI_KEY" \ -F 'prompt="Use the uploaded images in order: keep the boy wearing glasses and a blue T-shirt from [Image 1], add the corgi puppy from [Image 2], and use the lawn from [Image 3] as the setting. Create a 4-second 3D cartoon video with the boy and puppy sitting together, a slow camera push-in, soft daylight, and no extra characters."' \ -F 'model="doubao-seedance-2-0"' \ -F 'seconds="4"' \ -F 'size="720x1280"' \ -F 'input_reference=@"/path/to/seelite_ref_1.png"' \ -F 'input_reference=@"/path/to/seelite_ref_2.png"' \ -F 'input_reference=@"/path/to/seelite_ref_3.png"' - lang: Python label: Multiple reference images (Seedance 2.0) source: > import os import requests prompt = "Use the uploaded images in order: keep the boy wearing glasses and a blue T-shirt from [Image 1], add the corgi puppy from [Image 2], and use the lawn from [Image 3] as the setting. Create a 4-second 3D cartoon video with the boy and puppy sitting together, a slow camera push-in, soft daylight, and no extra characters." with open("/path/to/seelite_ref_1.png", "rb") as image_1, open("/path/to/seelite_ref_2.png", "rb") as image_2, open("/path/to/seelite_ref_3.png", "rb") as image_3: response = requests.post( "https://api.cometapi.com/v1/videos", headers={"Authorization": "Bearer " + os.environ["COMETAPI_KEY"]}, data={ "prompt": prompt, "model": "doubao-seedance-2-0", "seconds": "4", "size": "720x1280", }, files=[ ("input_reference", ("seelite_ref_1.png", image_1, "image/png")), ("input_reference", ("seelite_ref_2.png", image_2, "image/png")), ("input_reference", ("seelite_ref_3.png", image_3, "image/png")), ], timeout=30, ) response.raise_for_status() print(response.json()) - lang: JavaScript label: Multiple reference images (Seedance 2.0) source: > import { readFile } from "node:fs/promises"; const prompt = "Use the uploaded images in order: keep the boy wearing glasses and a blue T-shirt from [Image 1], add the corgi puppy from [Image 2], and use the lawn from [Image 3] as the setting. Create a 4-second 3D cartoon video with the boy and puppy sitting together, a slow camera push-in, soft daylight, and no extra characters."; const image1 = await readFile("/path/to/seelite_ref_1.png"); const image2 = await readFile("/path/to/seelite_ref_2.png"); const image3 = await readFile("/path/to/seelite_ref_3.png"); const form = new FormData(); form.append("prompt", prompt); form.append("model", "doubao-seedance-2-0"); form.append("seconds", "4"); form.append("size", "720x1280"); form.append("input_reference", new Blob([image1], { type: "image/png" }), "seelite_ref_1.png"); form.append("input_reference", new Blob([image2], { type: "image/png" }), "seelite_ref_2.png"); form.append("input_reference", new Blob([image3], { type: "image/png" }), "seelite_ref_3.png"); const response = await fetch("https://api.cometapi.com/v1/videos", { method: "POST", headers: { Authorization: `Bearer ${process.env.COMETAPI_KEY}` }, body: form, }); const result = await response.json(); console.log(result); - lang: Shell label: Text-to-video (exact 1920x1080) source: |- curl https://api.cometapi.com/v1/videos \ -H "Authorization: Bearer $COMETAPI_KEY" \ -F 'prompt="A slow cinematic camera push across a coastal landscape at sunrise"' \ -F 'model="doubao-seedance-1-5-pro"' \ -F 'seconds="4"' \ -F 'size="1920x1080"' - lang: Python label: Text-to-video (exact 1920x1080) source: | import os import requests response = requests.post( "https://api.cometapi.com/v1/videos", headers={"Authorization": "Bearer " + os.environ["COMETAPI_KEY"]}, data={ "prompt": "A slow cinematic camera push across a coastal landscape at sunrise", "model": "doubao-seedance-2-0", "seconds": "4", "size": "1920x1080", }, timeout=30, ) response.raise_for_status() print(response.json()) - lang: JavaScript label: Text-to-video (exact 1920x1080) source: > const form = new FormData(); form.append("prompt", "A slow cinematic camera push across a coastal landscape at sunrise"); form.append("model", "doubao-seedance-1-5-pro"); form.append("seconds", "4"); form.append("size", "1920x1080"); const response = await fetch("https://api.cometapi.com/v1/videos", { method: "POST", headers: { Authorization: `Bearer ${process.env.COMETAPI_KEY}` }, body: form, }); const result = await response.json(); console.log(result); components: schemas: CreateVideoResponse: type: object required: - id - object - model - status - progress - created_at properties: id: type: string description: Task id. Use it as the path parameter for GET /v1/videos/{id}. task_id: type: string description: Alias of id returned for compatibility. The value matches id. object: type: string description: Object type, always video. model: type: string description: Echo of the requested model id. status: type: string description: Initial task status. Newly created tasks are returned as queued. enum: - queued - in_progress - completed - failed - error progress: type: integer minimum: 0 maximum: 100 description: Completion percentage. 0 at creation. created_at: type: integer description: Task creation time as a Unix timestamp in seconds. additionalProperties: true ErrorResponse: description: >- Error body. The endpoint returns one of two shapes depending on where the validation fails. oneOf: - type: object properties: code: type: - string - 'null' description: Short error code such as invalid_request or InvalidParameter. message: type: string description: Human-readable error message. required: - message additionalProperties: true - type: object properties: error: type: object properties: code: type: - string - 'null' message: type: string type: type: string description: Error category, for example comet_api_error. required: - message - type additionalProperties: true required: - error additionalProperties: true securitySchemes: bearerAuth: type: http scheme: bearer description: Bearer token authentication. Use your CometAPI key. ````