文档 / 示例 / 图生视频

图生视频

将一张静态图片转换为短视频。该图片作为 reference_image 放入 metadata.content[]，并在 prompt 中用 @image1 引用它。

#流程

为你的图片获取一个 URL：通过 POST /v1/files 上传图片，或使用任何你已托管的公开 URL
提交 POST /v1/video/generations，将图片放入 metadata.content[]，并在 prompt 中用 @image1 引用
轮询 GET /v1/video/generations/{task_id} 直到状态为 succeeded，或设置 metadata.callback_url 等待 webhook 回调
从 url 字段下载视频

#cURL

terminalbash

# 1) 提交任务
curl https://relaydance.com/v1/video/generations \
  -H "Authorization: Bearer $RELAYDANCE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-1-5-pro-with-audio",
    "prompt": "@image1 comes alive: gentle camera pull-back, hair moving in a soft breeze",
    "seconds": "5",
    "metadata": {
      "ratio": "16:9",
      "resolution": "720p",
      "content": [
        {
          "type": "image_url",
          "image_url": { "url": "https://example.com/portrait.jpg" },
          "role": "reference_image"
        }
      ]
    }
  }'
# -> { "task_id": "...", "status": "submitted" }

# 2) 轮询（替换 TASK_ID）
curl https://relaydance.com/v1/video/generations/TASK_ID \
  -H "Authorization: Bearer $RELAYDANCE_API_KEY"

#Python（上传 + 提交 + 轮询）

image_to_video.pypython

import requests, os, time

BASE = "https://relaydance.com/v1"
HEADERS = {"Authorization": f"Bearer {os.environ['RELAYDANCE_API_KEY']}"}

# 1) 上传本地图片（若已有公开 URL 可跳过）
upload = requests.post(
    f"{BASE}/files",
    headers=HEADERS,
    files={"file": open("portrait.jpg", "rb")},
)
image_url = upload.json()["url"]

# 2) 提交视频任务
submit = requests.post(
    f"{BASE}/video/generations",
    headers={**HEADERS, "Content-Type": "application/json"},
    json={
        "model": "seedance-1-5-pro-with-audio",
        "prompt": "@image1 comes alive: gentle camera pull-back, soft natural light",
        "seconds": "5",
        "metadata": {
            "ratio": "16:9",
            "resolution": "720p",
            "content": [
                {
                    "type": "image_url",
                    "image_url": {"url": image_url},
                    "role": "reference_image",
                }
            ],
        },
    },
)
task_id = submit.json()["task_id"]

# 3) 轮询直到视频生成完成
while True:
    task = requests.get(f"{BASE}/video/generations/{task_id}", headers=HEADERS).json()
    if task["status"] == "succeeded":
        print("Video:", task["url"])
        break
    if task["status"] == "failed":
        raise RuntimeError(task)
    time.sleep(5)

#Node.js / TypeScript

imageToVideo.tstypescript

const BASE = "https://relaydance.com/v1";
const KEY = process.env.RELAYDANCE_API_KEY;

// 1) 使用托管图片 URL 提交任务
const submit = await fetch(`${BASE}/video/generations`, {
  method: "POST",
  headers: { Authorization: `Bearer ${KEY}`, "Content-Type": "application/json" },
  body: JSON.stringify({
    model: "seedance-1-5-pro-with-audio",
    prompt: "@image1 comes alive: gentle camera pull-back, soft natural light",
    seconds: "5",
    metadata: {
      ratio: "16:9",
      resolution: "720p",
      content: [
        {
          type: "image_url",
          image_url: { url: "https://example.com/portrait.jpg" },
          role: "reference_image",
        },
      ],
    },
  }),
});
const { task_id } = await submit.json();

// 2) 轮询
while (true) {
  const r = await fetch(`${BASE}/video/generations/${task_id}`, {
    headers: { Authorization: `Bearer ${KEY}` },
  });
  const task = await r.json();
  if (task.status === "succeeded") { console.log("Video:", task.url); break; }
  if (task.status === "failed") throw new Error(JSON.stringify(task));
  await new Promise((s) => setTimeout(s, 5000));
}

生产环境建议使用 webhook：在 metadata 内加上 callback_url，最终的任务 JSON 会 POST 给你，无需轮询循环。

#模型选择

模型	选择理由
`seedance-1-5-pro-with-audio`	图生视频，并生成配音
`seedance-1-5-pro-no-audio`	同系列，无声输出
`doubao-seedance-2-0-1080p`	最高保真度，多参考图融合
`happyhorse-1.0-i2v`	专用图生视频模型

每个请求最多 9 张参考图，按 content[] 的顺序以 @image1 到 @imageN 引用。参考图与首尾帧模式互斥。完整结构见视频生成。

←上传文件概览→