> ## Documentation Index
> Fetch the complete documentation index at: https://docs.aigc.it/llms.txt
> Use this file to discover all available pages before exploring further.

# 广告素材

> AI 生成广告投放所需的视频素材

## 场景介绍

广告投放需要大量视频素材进行 A/B 测试和多渠道投放。AI 图生视频可以快速生成多版本广告素材，提升投放效率。

## 适用场景

* **信息流广告**：抖音、快手、微信等平台的信息流视频广告
* **开屏广告**：App 开屏动态广告
* **Banner 视频**：网站动态 Banner
* **素材变体**：同一创意的多个版本用于 A/B 测试

## 推荐模型

| 场景    | 推荐模型                             | 原因          |
| ----- | -------------------------------- | ----------- |
| 信息流广告 | `wan2.6-i2v-flash-5s-720p`       | 快速生成，适合批量测试 |
| 高质量广告 | `wan2.6-i2v-5s-1080p`            | 更高质量，适合品牌广告 |
| 有声广告  | `wan2.6-i2v-flash-5s-720p-audio` | 自动配音，提升吸引力  |
| 长广告   | `wan2.6-i2v-15s-1080p`           | 完整故事叙述      |

## 示例：信息流广告

生成美妆产品的信息流广告视频：

```bash theme={null}
curl https://api.aigc.it/v1/videos \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <YOUR_TOKEN>" \
  -d '{
    "model": "wan2.6-i2v-flash-5s-720p",
    "prompt": "A beautiful woman applying lipstick, elegant movements, soft lighting, luxury beauty advertisement style, glamorous and sophisticated",
    "metadata": {
      "input": {
        "img_url": "https://example.com/beauty-ad.png"
      }
    }
  }'
```

## 示例：有声品牌广告

生成带有音效的品牌广告：

```bash theme={null}
curl https://api.aigc.it/v1/videos \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <YOUR_TOKEN>" \
  -d '{
    "model": "wan2.6-i2v-flash-5s-1080p-audio",
    "prompt": "Luxury car driving through city streets at night, neon lights reflecting on the car surface, cinematic movement, premium automotive advertisement",
    "metadata": {
      "input": {
        "img_url": "https://example.com/car-ad.png"
      }
    }
  }'
```

## 批量生成素材

使用脚本批量生成多个广告变体：

```python theme={null}
import requests

API_KEY = "<YOUR_API_KEY>"
BASE_URL = "https://api.aigc.it/v1"

prompts = [
    "Dynamic product showcase, energetic movement",
    "Elegant slow motion, luxury feel",
    "Fast-paced action, exciting atmosphere"
]

for i, prompt in enumerate(prompts):
    response = requests.post(
        f"{BASE_URL}/videos",
        headers={
            "Authorization": f"Bearer {API_KEY}",
            "Content-Type": "application/json"
        },
        json={
            "model": "wan2.6-i2v-flash-5s-720p",
            "prompt": prompt,
            "metadata": {
                "input": {
                    "img_url": "https://example.com/product.png"
                }
            }
        }
    )
    print(f"Variant {i+1}: {response.json()}")
```

## 最佳实践

<Tip>
  1. **多版本测试**：生成多个版本进行 A/B 测试
  2. **平台适配**：根据投放平台选择合适的分辨率和时长
  3. **有声优先**：有声视频通常有更高的转化率
</Tip>

## 相关文档

<CardGroup cols={2}>
  <Card title="万相 图生视频" icon="video" href="/zh/video-api/wan-i2v">
    查看完整的视频生成 API 文档
  </Card>

  <Card title="营销海报" icon="bullhorn" href="/zh/use-cases/marketing">
    了解营销图片生成场景
  </Card>
</CardGroup>