> ## 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.

# 创建 xAI 视频编辑

> 使用 POST /grok/v1/videos/edits 通过文本 Prompt 编辑源视频、保留运动效果，并获取用于异步轮询结果的 request_id。

使用此端点可通过文本指令编辑现有 MP4。与全新生成请求相比，输出会更贴近源片段的时序和构图。

## 发送请求之前

* 提供可访问的 `video.url`
* 保持源片段较短；xAI 官方建议将编辑时长限制在约 8.7 秒以内
* 使用聚焦的指令，描述一个明确的改动
* 保存返回的 `request_id`，因为编辑使用与生成相同的轮询流程

## 编辑流程

<Steps>
  <Step title="提交编辑请求">
    发送源视频 URL、编辑 Prompt 和 `model: grok-imagine-video`。
  </Step>

  <Step title="轮询最终结果">
    调用[获取 xAI 视频结果](./get-video-generation-results)，直到任务完成。
  </Step>

  <Step title="存储编辑后资源">
    下载已完成的输出，或将返回的 URL 转入你自己的存储流程。
  </Step>
</Steps>

## CometAPI 上的变化

xAI 将视频编辑文档描述为与生成相同的异步生命周期，只是将可选的源图像替换为源视频。CometAPI 保持了这一行为以及相同的轮询端点，因此你的编辑工作流仍然是“启动 -> 轮询 -> 下载”。


## OpenAPI

````yaml api/openapi/video/xai/post-video-edit.openapi.json POST /grok/v1/videos/edits
openapi: 3.1.0
info:
  title: Video Edit API
  version: 1.0.0
  description: >-
    Create an asynchronous xAI Grok video edit job from an existing source
    video.
servers:
  - url: https://api.cometapi.com
security:
  - bearerAuth: []
paths:
  /grok/v1/videos/edits:
    post:
      summary: Create an xAI video edit job
      description: >-
        Start a Grok video editing job from a source MP4 and a natural-language
        edit instruction. Poll the same query endpoint used by video generation.
      operationId: video_edit
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              required:
                - prompt
                - video
              properties:
                prompt:
                  type: string
                  description: Edit instruction describing the change you want.
                  example: Add snow to the scene.
                model:
                  type: string
                  description: xAI video model id.
                  default: grok-imagine-video
                  example: grok-imagine-video
                video:
                  type: object
                  required:
                    - url
                  properties:
                    url:
                      type: string
                      description: Reachable source MP4 URL.
                  description: >-
                    Source video to edit. xAI documents an input limit of about
                    8.7 seconds.
                output:
                  type: object
                  description: Optional output delivery configuration.
                user:
                  type: string
                  description: Optional end-user identifier.
              default:
                prompt: Add snow to the scene.
                model: grok-imagine-video
                video:
                  url: https://example.com/source.mp4
            examples:
              Edit request:
                summary: Edit request
                value:
                  prompt: Add snow to the scene.
                  video:
                    url: https://example.com/source.mp4
                  model: grok-imagine-video
      responses:
        '200':
          description: Request accepted.
          content:
            application/json:
              schema:
                type: object
                required:
                  - request_id
                properties:
                  request_id:
                    type: string
                    description: Deferred request id used for polling.
                example:
                  request_id: e55813f7-911f-cfa8-208c-9c8e693b4d38
              example:
                request_id: <request_id>
      x-codeSamples:
        - lang: Shell
          label: Default
          source: |
            curl https://api.cometapi.com/grok/v1/videos/edits \
              -H "Authorization: Bearer $COMETAPI_KEY" \
              -H "Content-Type: application/json" \
              -d '{
                "model": "grok-imagine-video",
                "prompt": "Add snow to the scene.",
                "video": {"url": "https://your-video-host/source.mp4"}
              }'
        - lang: Python
          label: Default
          source: |
            import os
            import requests

            response = requests.post(
                "https://api.cometapi.com/grok/v1/videos/edits",
                headers={"Authorization": "Bearer " + os.environ["COMETAPI_KEY"]},
                json={
                    "model": "grok-imagine-video",
                    "prompt": "Add snow to the scene.",
                    "video": {"url": "https://your-video-host/source.mp4"},
                },
            )

            task = response.json()
            print(task["request_id"])  # poll GET /grok/v1/videos/{request_id}
        - lang: JavaScript
          label: Default
          source: >
            const response = await
            fetch("https://api.cometapi.com/grok/v1/videos/edits", {
                method: "POST",
                headers: {
                    Authorization: `Bearer ${process.env.COMETAPI_KEY}`,
                    "Content-Type": "application/json",
                },
                body: JSON.stringify({
                    model: "grok-imagine-video",
                    prompt: "Add snow to the scene.",
                    video: { url: "https://your-video-host/source.mp4" },
                }),
            });


            const task = await response.json();

            console.log(task.request_id); // poll GET
            /grok/v1/videos/{request_id}
components:
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: Bearer token authentication. Use your CometAPI key.

````