查询与下载

五、查询任务状态

curl 'https://newapi.nskeven.uk/v1/video/generations/task_385412' \
  -H 'Authorization: Bearer sk-xxx'

返回示例

处理中

{
  "code": "success",
  "message": "",
  "data": {
    "id": 883,
    "task_id": "task_385412",
    "status": "IN_PROGRESS",
    "progress": "45%",
    "fail_reason": "",
    "result_url": "",
    "created_at": 1790224802,
    "finish_time": 0
  }
}

已完成

{
  "code": "success",
  "message": "",
  "data": {
    "id": 883,
    "task_id": "task_385412",
    "status": "SUCCESS",
    "progress": "100%",
    "fail_reason": "",
    "result_url": "https://newapi.nskeven.uk/v1/videos/task_385412/content",
    "created_at": 1790224802,
    "finish_time": 1790227431
  }
}

已失败

{
  "code": "success",
  "message": "",
  "data": {
    "id": 883,
    "task_id": "task_385412",
    "status": "FAILURE",
    "progress": "100%",
    "fail_reason": "生成被拒绝，内容安全审核未通过，修改提示词或素材后重试",
    "result_url": "",
    "created_at": 1790224802,
    "finish_time": 1790224804
  }
}

返回字段说明

状态说明

轮询建议

间隔 10～20 秒（sd-2-* 系列建议 ≥20 秒）。一般模型约 10 分钟内结束，超长时长约 30 分钟。超时未结束请保留 task_id 联系我们，不要重复提交。

失败原因

优先读取 data.fail_reason。若内容过于笼统，更具体的原因通常在 data.data 内部。

六、下载成片

curl -o out.mp4 'https://newapi.nskeven.uk/v1/videos/task_385412/content' \
  -H 'Authorization: Bearer sk-xxx'

返回视频文件数据流，直接写入文件即可。大文件请预留足够超时。

另有带下载头的等价端点，适合浏览器直接触发下载：

GET /v1/videos/{task_id}/download

关于 result_url

任务 SUCCESS 后，查询响应里的 data.result_url 恒为本平台地址，可直接用于下载：

https://newapi.nskeven.uk/v1/videos/{task_id}/content

完整调用示例（Python）

import time, requests

BASE = "https://newapi.nskeven.uk"
H = {"Authorization": "Bearer sk-xxx"}

tid = requests.post(f"{BASE}/v1/video/generations", headers=H, json={
    "model": "even-sd2.5-A",
    "prompt": "一只猫在雨夜霓虹街道奔跑，电影感",
    "seconds": "8", "resolution": "720p", "aspect_ratio": "9:16",
}).json()["id"]

while True:
    r = requests.get(f"{BASE}/v1/video/generations/{tid}", headers=H).json()["data"]
    print(r["status"], r["progress"])
    if r["status"] == "SUCCESS":
        break
    if r["status"] == "FAILURE":
        raise RuntimeError(f"生成失败：{r['fail_reason']}")
    time.sleep(15)

with requests.get(f"{BASE}/v1/videos/{tid}/content", headers=H, stream=True) as resp:
    resp.raise_for_status()
    with open("out.mp4", "wb") as f:
        for chunk in resp.iter_content(1 << 20):
            f.write(chunk)