如何存取 Veo 3 Pro API

步驟 1：註冊取得 API 金鑰

登入 cometapi.com。若您尚未成為我們的用戶，請先註冊。登入您的 CometAPI 控制台。取得介面存取憑證 API 金鑰。在個人中心的 API token 處點擊 “Add Token”，取得 token key：sk-xxxxx 並提交。

步驟 2：向 Veo 3 Pro API 發送請求

選擇 “\veo3-pro \” 端點發送 API 請求，並設定請求體。請求方法與請求體可於我們網站的 API 文件取得。我們的網站也提供 Apifox 測試以供方便。將 <YOUR_API_KEY> 替換為您帳戶中的實際 CometAPI 金鑰。基礎 URL 是 Veo3 異步生成(https://api.cometapi.com/v1/videos)。

將您的問題或請求插入 content 欄位——模型將對此作出回應。處理 API 回應以取得生成的答案。

步驟 3：擷取並驗證結果

處理 API 回應以取得生成的答案。處理完成後，API 會返回任務狀態與輸出資料。

如需進一步了解 Veo3，請參見 Veo3 頁面。

Veo 3 Pro 的範例程式碼和 API

存取完整的範例程式碼和 API 資源，以簡化您的 Veo 3 Pro 整合流程。我們詳盡的文件提供逐步指引，協助您在專案中充分發揮 Veo 3 Pro 的潛力。

Python
JavaScript
Curl

import os
import time
import requests

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com/veo/v1/video"

# Create video generation task
create_response = requests.post(
    f"{BASE_URL}/create",
    headers={
        "Authorization": COMETAPI_KEY,
        "Content-Type": "application/json",
    },
    json={
        "prompt": "An orange cat flying in the blue sky with white clouds, sunlight pouring onto its fur, creating a beautiful and dreamlike scene",
        "model": "veo3.1-pro",
        "enhance_prompt": True,
    },
)

task = create_response.json()
task_id = task["id"]
print(f"Task created: {task_id}")
print(f"Status: {task['status']}")

# Poll until video is ready
while True:
    query_response = requests.get(
        f"{BASE_URL}/query/{task_id}",
        headers={
            "Authorization": f"Bearer {COMETAPI_KEY}",
        },
    )

    result = query_response.json()
    status = result["data"]["status"]
    progress = result["data"].get("progress", "")

    print(f"Checking status... {status} {progress}")

    if status == "SUCCESS" or result["data"]["data"]["status"] == "completed":
        video_url = result["data"]["data"]["video_url"]
        print(f"
Video URL: {video_url}")
        break
    elif status == "FAILED":
        print(f"Failed: {result['data'].get('fail_reason', 'Unknown error')}")
        break

    time.sleep(10)

Python Code Example

import os
import time
import requests

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com/veo/v1/video"

# Create video generation task
create_response = requests.post(
    f"{BASE_URL}/create",
    headers={
        "Authorization": COMETAPI_KEY,
        "Content-Type": "application/json",
    },
    json={
        "prompt": "An orange cat flying in the blue sky with white clouds, sunlight pouring onto its fur, creating a beautiful and dreamlike scene",
        "model": "veo3.1-pro",
        "enhance_prompt": True,
    },
)

task = create_response.json()
task_id = task["id"]
print(f"Task created: {task_id}")
print(f"Status: {task['status']}")

# Poll until video is ready
while True:
    query_response = requests.get(
        f"{BASE_URL}/query/{task_id}",
        headers={
            "Authorization": f"Bearer {COMETAPI_KEY}",
        },
    )

    result = query_response.json()
    status = result["data"]["status"]
    progress = result["data"].get("progress", "")

    print(f"Checking status... {status} {progress}")

    if status == "SUCCESS" or result["data"]["data"]["status"] == "completed":
        video_url = result["data"]["data"]["video_url"]
        print(f"\nVideo URL: {video_url}")
        break
    elif status == "FAILED":
        print(f"Failed: {result['data'].get('fail_reason', 'Unknown error')}")
        break

    time.sleep(10)

JavaScript Code Example

// Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
const COMETAPI_KEY = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const BASE_URL = "https://api.cometapi.com/veo/v1/video";

async function sleep(ms) {
  return new Promise((resolve) => setTimeout(resolve, ms));
}

async function main() {
  // Create video generation task
  const createResponse = await fetch(`${BASE_URL}/create`, {
    method: "POST",
    headers: {
      Authorization: COMETAPI_KEY,
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      prompt:
        "An orange cat flying in the blue sky with white clouds, sunlight pouring onto its fur, creating a beautiful and dreamlike scene",
      model: "veo3.1-pro",
      enhance_prompt: true,
    }),
  });

  const task = await createResponse.json();
  const taskId = task.id;
  console.log(`Task created: ${taskId}`);
  console.log(`Status: ${task.status}`);

  // Poll until video is ready
  while (true) {
    const queryResponse = await fetch(`${BASE_URL}/query/${taskId}`, {
      headers: {
        Authorization: `Bearer ${COMETAPI_KEY}`,
      },
    });

    const result = await queryResponse.json();
    const status = result.data.status;
    const progress = result.data.progress || "";

    console.log(`Checking status... ${status} ${progress}`);

    if (status === "SUCCESS" || result.data.data.status === "completed") {
      const videoUrl = result.data.data.video_url;
      console.log(`\nVideo URL: ${videoUrl}`);
      break;
    } else if (status === "FAILED") {
      console.log(`Failed: ${result.data.fail_reason || "Unknown error"}`);
      break;
    }

    await sleep(10000);
  }
}

main();

Curl Code Example

#!/bin/bash
# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here

BASE_URL="https://api.cometapi.com/veo/v1/video"

# Create video generation task
response=$(curl -s -X POST "${BASE_URL}/create" \
  -H "Authorization: $COMETAPI_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "An orange cat flying in the blue sky with white clouds, sunlight pouring onto its fur, creating a beautiful and dreamlike scene",
    "model": "veo3.1-pro",
    "enhance_prompt": true
  }')

task_id=$(echo "$response" | jq -r '.id')
status=$(echo "$response" | jq -r '.status')

echo "Task created: ${task_id}"
echo "Status: ${status}"

# Poll until video is ready
while true; do
  query_response=$(curl -s -X GET "${BASE_URL}/query/${task_id}" \
    -H "Authorization: Bearer $COMETAPI_KEY")

  status=$(echo "$query_response" | jq -r '.data.status')
  progress=$(echo "$query_response" | jq -r '.data.progress // ""')
  completed=$(echo "$query_response" | jq -r '.data.data.status')

  echo "Checking status... ${status} ${progress}"

  if [ "$status" = "SUCCESS" ] || [ "$completed" = "completed" ]; then
    video_url=$(echo "$query_response" | jq -r '.data.data.video_url')
    echo ""
    echo "Video URL: ${video_url}"
    break
  elif [ "$status" = "FAILED" ]; then
    fail_reason=$(echo "$query_response" | jq -r '.data.fail_reason // "Unknown error"')
    echo "Failed: ${fail_reason}"
    break
  fi

  sleep 10
done

Veo 3 Pro的版本

Veo 3 Pro擁有多個快照的原因可能包括：更新後輸出結果存在差異需保留舊版快照以確保一致性、為開發者提供適應與遷移的過渡期，以及不同快照對應全球或區域端點以優化使用者體驗等潛在因素。各版本間的具體差異請參閱官方文件說明。

veo3-pro
veo3-pro-frames	veo3-frames 模型特別針對影格序列生成進行優化。veo3-frames 模型特別針對影格序列生成進行優化，並包含一個示意圖，支援首尾影格。