douyin-video

Douyin watermark-free video download and copy extraction

Obtain a watermark-free video download link from a Douyin share link, download the video, and use speech recognition to extract the copy from the video, automatically saving it to a file.

Feature Overview

• Obtain Download Link: Parse a watermark-free video direct download address from a Douyin share link (no API key required)
• Download Video: Download the watermark-free video to a specified local directory
• Extract Copy: Extract text content from the video via speech recognition (requires a SiliconFlow API key)
• Automatic Save: Each video's copy is automatically saved to a separate folder (with the video ID as the folder name)

Environment Requirements

Dependency Installation

pip install requests ffmpeg-python

System Requirements

• FFmpeg must be installed on the system (for audio and video processing)
• macOS: brew install ffmpeg
• Ubuntu: apt install ffmpeg

API Key Configuration (only required for copy extraction)

The copy extraction function uses the SiliconFlow API, and an environment variable needs to be set:

export API_KEY="your-siliconflow-api-key"

Obtain API key: https://cloud.siliconflow.cn/

Usage

Method 1: Use the script (recommended)

# 获取视频信息和下载链接 (无需 API 密钥)
python douyin_downloader.py --link "抖音分享链接" --action info

# 下载视频到指定目录
python douyin_downloader.py --link "抖音分享链接" --action download --output ./videos

# 提取视频文案并保存到文件 (需要 API_KEY 环境变量)
python douyin_downloader.py --link "抖音分享链接" --action extract --output ./output

# 提取文案并同时保存视频
python douyin_downloader.py --link "抖音分享链接" --action extract --output ./output --save-video

# 安静模式 (减少输出)
python douyin_downloader.py --link "抖音分享链接" --action extract --output ./output --quiet

Output Directory Structure

After extracting the copy, each video is saved to a separate folder:

output/
├── 7600361826030865707/      # 视频ID为文件夹名
│   └── transcript.md         # Markdown 格式文案文件
├── 7581044356631612699/
│   ├── transcript.md
│   └── 7581044356631612699.mp4  # 使用 --save-video 时保存
└── ...

Markdown Copy Format

# 视频标题

| 属性 | 值 |
|------|-----|
| 视频ID | `7600361826030865707` |
| 提取时间 | 2026-01-30 14:19:00 |
| 下载链接 | [点击下载](url) |

---

## 文案内容

(语音识别的文字内容)

Method 2: Call in Python code

from douyin_downloader import get_video_info, download_video, extract_text

# 获取视频信息
info = get_video_info("抖音分享链接")
print(f"视频ID: {info['video_id']}")
print(f"标题: {info['title']}")
print(f"下载链接: {info['url']}")

# 下载视频
video_path = download_video("抖音分享链接", output_dir="./videos")

# 提取文案并保存到文件
result = extract_text("抖音分享链接", output_dir="./output")
print(f"文案已保存到: {result['output_path']}")
print(result['text'])

Workflow

Obtain Video Information

1. Parse the Douyin share link and extract the real video URL
2. Simulate a mobile client request to obtain page data
3. Extract the watermark-free video address from the page JSON data
4. Return the video ID, title, and download link

Extract Video Copy

1. Parse the share link to obtain video information
2. Download the video to a temporary directory
3. Use FFmpeg to extract audio from the video (MP3 format)
4. Call the SiliconFlow SenseVoice API for speech recognition
5. Clean up temporary files and return the recognized text

FAQ

Unable to Parse Link

• Ensure the link is a valid Douyin share link
• The link format is usually https://v.douyin.com/xxxxx/ or a complete Douyin video URL

Copy Extraction Failed

• Check whether the API_KEY environment variable has been set
• Ensure the API key is valid and has sufficient quota
• Ensure FFmpeg is correctly installed

Slow Download Speed

• This depends on network conditions and video size
• The script will display download progress

Notes

• This tool is for learning and research purposes only
• Relevant laws and regulations must be complied with when using it
• Do not use it for any purpose that infringes copyright or violates the law