返回模型调用 API 参考
POST/kling/v1/videos/omni-video

Omni-Video

API Key application/json

请求示例

curl -X POST https://top.qingyuntop.ai/kling/v1/videos/omni-video \
  -H "Authorization: Bearer ${QINGYUN_TOP_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "kling-video-o1",
    "prompt": "生成一个海边有一个人跳舞的视频",
    "negative_prompt": "",
    // "cfg_scale": 0.5,
    "mode": "std",
    // "sound": "off",
    // "image_list": [
    //     {
    //         "image_url": "https://h2.inkwai.com/bs2/upload-ylab-stunt/se/ai_portal_queue_mmu_image_upscale_aiweb/3214b798-e1b4-4b00-b7af-72b5b0417420_raw_image_0.jpg",
    //         "type": "first_frame"
    //     }
    // ],
    // "video_list": [
    //     {
    //         "video_url": "",
    //         "refer_type": "base",
    //         "keep_original_sound": "yes"
    //     }
    // ],
    "element_list": [
        {
            "element_id": 835266081714884677
        }
    ],
    "aspect_ratio": "16:9",
    "duration": "5",
    "callback_url": "",
    "external_task_id": ""
}'

请求参数

参数位置类型必填说明示例
Content-TypeHeaderstring-application/json
AuthorizationHeaderstring-Bearer {{YOUR_API_KEY}}
model_nameJSON bodystring模型名称 枚举值:kling-video-o1 kling-v3-omni-
multi_shotJSON bodyboolean是否生成多镜头视频 当前参数为true时,prompt参数无效 当前参数为false时,shot_type参数及multi_prompt参数无效-
shot_typeJSON bodystring分镜方式 枚举值:customize 当multi_shot参数为true时,当前参数必填-
promptJSON bodystring文本提示词,可包含正向描述和负向描述 可将提示词模板化来满足不同的视频生成需求 不能超过2500个字符 当multi_shot参数为false,当前参数不得为空-
multi_promptJSON bodyarray<object>各分镜信息,如提示词、时长等 ● 通过index、prompt、duration参数定义分镜序号及相应提示词和时长,其中: ○ 最多支持6个分镜,最小支持1个分镜 ○ 每个分镜相关内容的最大长度不超过512 ○ 每个分镜的时长不大于当前任务的总时长,不小于1 ○ 所有分镜的时长之和等于当前任务的总时长-
multi_prompt[].indexJSON bodynumber镜序号-
multi_prompt[].promptJSON bodystring提示词-
multi_prompt[].durationJSON bodystring时长-
soundJSON bodystring生成视频时是否同时生成声音 枚举值:on,off-
image_listJSON bodyarray<object>参考图列表 包括主体、场景、风格等参考图片,也可作为首帧或尾帧生成视频;当作为首帧或尾帧生成视频时: 通过type参数来定义图片是否为首尾帧:first_frame为首帧,end_frame为尾帧 暂时不支持仅尾帧,即有尾帧图时必须有首帧图 首帧或首尾帧生视频时,不能使用视频编辑功能-
image_list[].image_urlJSON bodystring--
image_list[].typeJSON bodystring--
video_listJSON bodyarray<object>参考视频,通过URL方式获取 可作为特征参考视频,也可作为待编辑视频,默认为待编辑视频;可选择性保留视频原声 通过refer_type参数区分参考视频类型:feature为特征参考视频,base为待编辑视频 参考视频为待编辑视频时,不能定义视频首尾帧 通过keep_original_sound参数选择是否保留视频原声,yes为保留,no为不保留;当前参数对特征参考视频(feature)也生效-
video_list[].video_urlJSON bodystring--
video_list[].refer_typeJSON bodystring--
video_list[].keep_original_soundJSON bodystring--
element_listJSON bodyarray<object>主体-
element_list[].element_idJSON bodystring主体id-
modeJSON bodystring生成视频的模式 枚举值:std,pro 其中std:标准模式(标准),基础模式,性价比高 其中pro:专家模式(高品质),高表现模式,生成视频质量更佳-
aspect_ratioJSON bodystring生成视频的画面纵横比(宽:高) 枚举值:16:9, 9:16, 1:1 未使用首帧参考或视频编辑功能时,当前参数必填-
durationJSON bodystring生成视频时长,单位s 枚举值:3,4,5,6,7,8,9,10,其中: 使用文生视频、首帧图生视频时,仅支持5和10s 使用视频编辑功能(“refer_type”:“base”)时,输出结果与传入视频时长相同,此时当前参数无效;此时,按输入视频时长四舍五入取整计量计费 kling-v3-omni 支持 ● 枚举值:3,4,5,6,7,8,9,10,11,12,13,14,15,其中: 使用视频编辑功能("refer_type":"base")时,输出结果与传入视频时长相同,此时当前参数无效;此时,按输入视频时长四舍五入取整计量计费-
watermark_infoJSON bodyobject是否同时生成含水印的结果 ● 通过enabled参数定义,用key:value承载,如下:: "watermark_info": { "enabled": boolean // true 为生成,false 为不生成 }-
watermark_info.enabledJSON bodyboolean--
callback_urlJSON bodystring--
external_task_idJSON bodystring--

返回响应 (200)

此部分没有定义参数。

来源原始接口定义(用于追溯)

以下内容来自 Apifox 快照,可能包含兼容层字段或旧示例;公开调用请以本页的协议说明和当前模型列表为准。

{
  "id": 495899786,
  "name": "Omni-Video",
  "type": "http",
  "serverId": "",
  "preProcessors": [],
  "postProcessors": [],
  "inheritPreProcessors": {},
  "inheritPostProcessors": {},
  "description": "",
  "operationId": "",
  "sourceUrl": "",
  "method": "post",
  "path": "/kling/v1/videos/omni-video",
  "tags": [],
  "status": 1,
  "requestBody": {
    "type": "application/json",
    "parameters": [],
    "jsonSchema": {
      "type": "object",
      "properties": {
        "model_name": {
          "type": "string",
          "description": "模型名称 枚举值:kling-video-o1  kling-v3-omni"
        },
        "prompt": {
          "type": "string",
          "description": "文本提示词,可包含正向描述和负向描述\n可将提示词模板化来满足不同的视频生成需求\n不能超过2500个字符\n当multi_shot参数为false,当前参数不得为空"
        },
        "image_list": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "image_url": {
                "type": "string"
              },
              "type": {
                "type": "string"
              }
            },
            "x-apifox-orders": [
              "image_url",
              "type"
            ]
          },
          "description": "参考图列表\n包括主体、场景、风格等参考图片,也可作为首帧或尾帧生成视频;当作为首帧或尾帧生成视频时:\n通过type参数来定义图片是否为首尾帧:first_frame为首帧,end_frame为尾帧\n暂时不支持仅尾帧,即有尾帧图时必须有首帧图\n首帧或首尾帧生视频时,不能使用视频编辑功能"
        },
        "video_list": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "video_url": {
                "type": "string"
              },
              "refer_type": {
                "type": "string"
              },
              "keep_original_sound": {
                "type": "string"
              }
            },
            "x-apifox-orders": [
              "video_url",
              "refer_type",
              "keep_original_sound"
            ]
          },
          "description": "参考视频,通过URL方式获取\n可作为特征参考视频,也可作为待编辑视频,默认为待编辑视频;可选择性保留视频原声\n通过refer_type参数区分参考视频类型:feature为特征参考视频,base为待编辑视频\n参考视频为待编辑视频时,不能定义视频首尾帧\n通过keep_original_sound参数选择是否保留视频原声,yes为保留,no为不保留;当前参数对特征参考视频(feature)也生效"
        },
        "mode": {
          "type": "string",
          "description": "生成视频的模式\n枚举值:std,pro\n其中std:标准模式(标准),基础模式,性价比高\n其中pro:专家模式(高品质),高表现模式,生成视频质量更佳"
        },
        "aspect_ratio": {
          "type": "string",
          "description": "生成视频的画面纵横比(宽:高)\n枚举值:16:9, 9:16, 1:1\n未使用首帧参考或视频编辑功能时,当前参数必填"
        },
        "duration": {
          "type": "string",
          "description": "生成视频时长,单位s\n枚举值:3,4,5,6,7,8,9,10,其中:\n使用文生视频、首帧图生视频时,仅支持5和10s\n使用视频编辑功能(“refer_type”:“base”)时,输出结果与传入视频时长相同,此时当前参数无效;此时,按输入视频时长四舍五入取整计量计费\nkling-v3-omni  支持\n● 枚举值:3,4,5,6,7,8,9,10,11,12,13,14,15,其中:\n使用视频编辑功能(\"refer_type\":\"base\")时,输出结果与传入视频时长相同,此时当前参数无效;此时,按输入视频时长四舍五入取整计量计费"
        },
        "callback_url": {
          "type": "string"
        },
        "external_task_id": {
          "type": "string"
        },
        "element_list": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "element_id": {
                "type": "string",
                "description": "主体id"
              }
            },
            "x-apifox-orders": [
              "element_id"
            ]
          },
          "description": "主体"
        },
        "multi_shot": {
          "type": "boolean",
          "description": "是否生成多镜头视频\n当前参数为true时,prompt参数无效\n当前参数为false时,shot_type参数及multi_prompt参数无效"
        },
        "shot_type": {
          "type": "string",
          "description": "分镜方式\n枚举值:customize\n当multi_shot参数为true时,当前参数必填"
        },
        "multi_prompt": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "index": {
                "type": "number",
                "description": "镜序号"
              },
              "prompt": {
                "type": "string",
                "description": "提示词"
              },
              "duration": {
                "type": "string",
                "description": "时长"
              }
            },
            "x-apifox-orders": [
              "index",
              "prompt",
              "duration"
            ]
          },
          "description": "各分镜信息,如提示词、时长等\n● 通过index、prompt、duration参数定义分镜序号及相应提示词和时长,其中:\n\t○ 最多支持6个分镜,最小支持1个分镜\n\t○ 每个分镜相关内容的最大长度不超过512\n\t○ 每个分镜的时长不大于当前任务的总时长,不小于1\n\t○ 所有分镜的时长之和等于当前任务的总时长"
        },
        "watermark_info": {
          "type": "object",
          "properties": {
            "enabled": {
              "type": "boolean"
            }
          },
          "x-apifox-orders": [
            "enabled"
          ],
          "description": "是否同时生成含水印的结果\n● 通过enabled参数定义,用key:value承载,如下::\n\"watermark_info\": {\n \t\"enabled\": boolean // true 为生成,false 为不生成\n}"
        },
        "sound": {
          "type": "string",
          "description": "生成视频时是否同时生成声音\n枚举值:on,off"
        }
      },
      "required": [
        "model_name",
        "mode",
        "duration",
        "multi_shot"
      ],
      "x-apifox-orders": [
        "model_name",
        "multi_shot",
        "shot_type",
        "prompt",
        "multi_prompt",
        "sound",
        "image_list",
        "video_list",
        "element_list",
        "mode",
        "aspect_ratio",
        "duration",
        "watermark_info",
        "callback_url",
        "external_task_id"
      ]
    },
    "examples": [
      {
        "mediaType": "application/json",
        "value": "{\r\n    \"model_name\": \"kling-video-o1\",\r\n    \"prompt\": \"生成一个海边有一个人跳舞的视频\",\r\n    \"negative_prompt\": \"\",\r\n    // \"cfg_scale\": 0.5,\r\n    \"mode\": \"std\",\r\n    // \"sound\": \"off\",\r\n    // \"image_list\": [\r\n    //     {\r\n    //         \"image_url\": \"https://h2.inkwai.com/bs2/upload-ylab-stunt/se/ai_portal_queue_mmu_image_upscale_aiweb/3214b798-e1b4-4b00-b7af-72b5b0417420_raw_image_0.jpg\",\r\n    //         \"type\": \"first_frame\"\r\n    //     }\r\n    // ],\r\n    // \"video_list\": [\r\n    //     {\r\n    //         \"video_url\": \"\",\r\n    //         \"refer_type\": \"base\",\r\n    //         \"keep_original_sound\": \"yes\"\r\n    //     }\r\n    // ],\r\n    \"element_list\": [\r\n        {\r\n            \"element_id\": 835266081714884677\r\n        }\r\n    ],\r\n    \"aspect_ratio\": \"16:9\",\r\n    \"duration\": \"5\",\r\n    \"callback_url\": \"\",\r\n    \"external_task_id\": \"\"\r\n}"
      }
    ],
    "mediaType": "",
    "oasExtensions": "",
    "required": true,
    "additionalContentTypes": []
  },
  "parameters": {
    "path": [],
    "query": [],
    "cookie": [],
    "header": [
      {
        "id": "sdTrFfapNw",
        "name": "Content-Type",
        "required": false,
        "enable": true,
        "description": "",
        "type": "string",
        "sampleValue": "application/json"
      },
      {
        "id": "ymxk8tgbx1",
        "name": "Authorization",
        "required": false,
        "enable": true,
        "description": "",
        "type": "string",
        "sampleValue": "Bearer {{YOUR_API_KEY}}"
      }
    ]
  },
  "commonParameters": {
    "query": [],
    "body": [],
    "cookie": [],
    "header": []
  },
  "auth": {},
  "responses": [
    {
      "id": 170970450,
      "name": "成功",
      "code": 200,
      "contentType": "json",
      "jsonSchema": {
        "type": "object",
        "properties": {},
        "x-apifox-orders": []
      },
      "itemSchema": {},
      "description": "",
      "mediaType": "",
      "headers": []
    }
  ],
  "responseExamples": [],
  "codeSamples": [],
  "projectId": 8662722,
  "moduleId": 8310462,
  "folderId": 92436141,
  "ordering": 0,
  "responsibleId": 0,
  "commonResponseStatus": {},
  "advancedSettings": {
    "disabledSystemHeaders": {}
  },
  "customApiFields": {},
  "oasExtensions": "",
  "mockScript": {},
  "createdAt": "2026-08-02T03:53:34.000Z",
  "updatedAt": "2026-08-06T12:15:55.000Z",
  "creatorId": 2863449,
  "editorId": 2863449,
  "creatorUserId": 2863449,
  "editorUserId": 2863449,
  "responseChildren": [],
  "visibility": "INHERITED",
  "securityScheme": {},
  "callbacks": ""
}