> ## 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. # 建立 Runway 角色表演 > 使用 CometAPI 搭配 Runway Control a character API，從參考影片啟動角色表演任務，驅動臉部表情與身體動作。 使用這個端點可將參考片段中的表演轉移到角色圖片或角色影片上。 ## 呼叫前準備 * 傳送必要的 `X-Runway-Version` 標頭，例如 `2024-11-06` * 使用 `model: act_two` * 提供一個 `character` 資產與一個 `reference` 資產，且都必須是可存取的 HTTPS URL * 在調整風格前，先以 `bodyControl: true` 和適中的 `expressionIntensity` 開始 ## 輸出比例 | 設定 | 支援的值 | 預設起始值 | 邊界行為 | | ------- | ------------------------------------------------------ | ---------- | --------------------- | | `ratio` | `1280:720`、`720:1280`，或所選 Runway model/version 接受的其他比例 | `1280:720` | 讓角色與參考資產盡量接近所要求的畫面比例。 | ## 任務流程 使用你的角色資產、參考資產與動作設定建立任務。 保留回傳的 `id`，因為後續輪詢會用到它。 持續使用[取得 Runway 任務](./runway-to-get-task-details)，直到任務進入終態，且提供者中繼資料包含完成的輸出。 ## 輸入角色 * `character` 是你想要製作動畫的主體 * `reference` 是提供臉部表演或身體動作的來源片段 * `bodyControl` 決定動作轉移是否除了表情轉移外，也包含身體動作 * `contentModeration` 會將提供者的審核門檻套用到此請求 ## OpenAPI ````yaml api/openapi/video/runway/official-format/post-control-a-character.openapi.json POST /runwayml/v1/character_performance openapi: 3.1.0 info: title: Control a Character API version: 1.0.0 description: >- Create a Runway character-performance task through the official-format CometAPI route. servers: - url: https://api.cometapi.com security: - bearerAuth: [] paths: /runwayml/v1/character_performance: post: summary: Create a Runway character-performance task description: >- Submit a character asset plus a reference performance clip and receive a task id for later polling. operationId: control_a_character parameters: - name: X-Runway-Version in: header required: true description: Runway version header, for example `2024-11-06`. schema: type: string requestBody: required: true content: application/json: schema: type: object required: - character - reference - bodyControl - expressionIntensity - seed - model - ratio - contentModeration properties: character: type: object description: >- The character to control. You can either provide a video or an image. required: - type - uri properties: type: type: string uri: type: string default: type: video uri: https://your-video-host/source.mp4 reference: type: object description: >- Reference performance clip. Provide `type` (`video`) and `uri` (HTTPS URL). required: - type - uri properties: type: type: string uri: type: string default: type: video uri: https://your-video-host/source.mp4 bodyControl: type: boolean description: Enable body control. default: true expressionIntensity: type: integer description: >- Intensity of the expression transfer. Higher values produce more exaggerated expressions. default: 1 seed: type: integer description: Random seed for reproducible results. default: 1 model: type: string description: The model variant to use. Must be act_two. default: act_two ratio: type: string description: >- Requested output frame shape, for example 1280:720 or 720:1280. Keep the character and reference assets close to the requested frame shape. default: '1280:720' contentModeration: type: object required: - publicFigureThreshold properties: publicFigureThreshold: type: string description: >- Threshold for public-figure detection. Use `auto` for default behavior. default: publicFigureThreshold: auto default: character: type: video uri: https://your-video-host/source.mp4 reference: type: video uri: https://your-video-host/source.mp4 bodyControl: true expressionIntensity: 1 seed: 1 model: act_two ratio: '1280:720' contentModeration: publicFigureThreshold: auto examples: Default: summary: Default value: character: type: video uri: https://your-video-host/source.mp4 reference: type: video uri: https://your-video-host/source.mp4 bodyControl: true expressionIntensity: 3 seed: 4294967295 model: act_two ratio: '1280:720' contentModeration: publicFigureThreshold: auto responses: '200': description: Task accepted. content: application/json: schema: type: object required: - id properties: id: type: string example: id: 5b91e444-c232-deaf-a4d7-8aef3e57336b x-codeSamples: - lang: Shell label: Default source: | curl https://api.cometapi.com/runwayml/v1/character_performance \ -H "Authorization: Bearer $COMETAPI_KEY" \ -H "Content-Type: application/json" \ -d '{ "character": { "type": "video", "uri": "https://your-video-host/character.mp4" }, "reference": { "type": "video", "uri": "https://your-video-host/performance.mp4" }, "model": "act_two", "ratio": "1280:720" }' - lang: Python label: Default source: > import os import requests response = requests.post( "https://api.cometapi.com/runwayml/v1/character_performance", headers={"Authorization": "Bearer " + os.environ["COMETAPI_KEY"]}, json={ "character": { "type": "video", "uri": "https://your-video-host/character.mp4" }, "reference": { "type": "video", "uri": "https://your-video-host/performance.mp4" }, "model": "act_two", "ratio": "1280:720" }, ) task = response.json() print(task["id"]) # poll GET /runwayml/v1/tasks/{id} until the output array appears - lang: JavaScript label: Default source: > const response = await fetch("https://api.cometapi.com/runwayml/v1/character_performance", { method: "POST", headers: { Authorization: `Bearer ${process.env.COMETAPI_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ "character": { "type": "video", "uri": "https://your-video-host/character.mp4" }, "reference": { "type": "video", "uri": "https://your-video-host/performance.mp4" }, "model": "act_two", "ratio": "1280:720" }), }); const task = await response.json(); console.log(task.id); // poll GET /runwayml/v1/tasks/{id} until the output array appears components: securitySchemes: bearerAuth: type: http scheme: bearer description: Bearer token authentication. Use your CometAPI key. ````