video-analyzer

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Video Analyzer

通过视觉/视频大模型分析视频内容，支持本地视频文件和互联网视频。

Analyze video content using visual/video large models, supporting local video files and online videos.

使用场景

Use Cases

用户要求分析、理解或描述一段视频
用户提供视频文件路径或 URL，希望了解视频内容
用户需要对视频进行问答

Users request to analyze, understand, or describe a video
Users provide a video file path or URL and want to know the video content
Users need to ask questions about the video

配置

Configuration

环境变量

Environment Variables

根据使用的模型设置对应的 API Key 环境变量：

bash

undefined

Set the corresponding API Key environment variables based on the model used:

bash

undefined

火山引擎（豆包）

VolcEngine (Doubao)

export ARK_API_KEY="your-api-key"

OpenAI

export OPENAI_API_KEY="your-api-key"

undefined

export OPENAI_API_KEY="your-api-key"

undefined

模型配置

Model Configuration

编辑

scripts/models.json

添加或修改模型配置。每个模型需要：

```
base_url
```
— API 地址
```
api_key_env
```
— 读取 API Key 的环境变量名
```
model
```
— 模型 ID
```
api_type
```
—
```
responses
```
或
```
chat_completions
```
```
supports_video
```
— 是否支持原生视频输入

Edit

scripts/models.json

to add or modify model configurations. Each model requires:

```
base_url
```
— API endpoint
```
api_key_env
```
— Name of the environment variable for reading the API Key
```
model
```
— Model ID
```
api_type
```
—
```
responses
```
or
```
chat_completions
```
```
supports_video
```
— Whether native video input is supported

工作流程

Workflow

确认视频来源：获取用户提供的视频路径或 URL。
确认分析需求：明确用户想了解什么（如概括内容、回答问题、描述场景等）。如果
```
$ARGUMENTS
```
非空，将其作为分析提示词。
选择模型：默认使用
```
models.json
```
中的
```
default_model
```
，用户也可指定。
执行分析：运行脚本（在
```
scripts/
```
目录下执行）：
bash
```
uv run analyze.py --video <视频路径或URL> --prompt "<分析提示词>"
```
可选参数：
- ```
--model <名称>
```
  — 指定模型（对应 models.json 中的 key）
- ```
--frames <数量>
```
  — 抽帧数量（默认 10）
- ```
--max-size <像素>
```
  — 帧最大边长（默认 720）
展示结果：将模型返回的分析结果展示给用户。

Confirm Video Source: Obtain the video path or URL provided by the user.
Confirm Analysis Requirements: Clarify what the user wants to know (e.g., content summary, question answering, scene description, etc.). If
```
$ARGUMENTS
```
is not empty, use it as the analysis prompt.
Select Model: By default, use the
```
default_model
```
in
```
models.json
```
; users can also specify a model.
Execute Analysis: Run the script (execute in the
```
scripts/
```
directory):
bash
```
uv run analyze.py --video <video path or URL> --prompt "<analysis prompt>"
```
Optional parameters:
- ```
--model <name>
```
  — Specify a model (corresponds to the key in models.json)
- ```
--frames <number>
```
  — Number of frames to extract (default: 10)
- ```
--max-size <pixels>
```
  — Maximum side length of frames (default: 720)
Display Results: Present the analysis results returned by the model to the user.

CLI 参考

CLI Reference

bash

undefined

bash

undefined

本地视频

Local video

uv run analyze.py --video /path/to/video.mp4 --prompt "描述视频内容"

uv run analyze.py --video /path/to/video.mp4 --prompt "Describe the video content"

互联网直接视频 URL

Direct online video URL

uv run analyze.py --video https://example.com/video.mp4 --prompt "分析视频"

uv run analyze.py --video https://example.com/video.mp4 --prompt "Analyze the video"

视频站点 URL（YouTube、Bilibili 等）

Video website URL (YouTube, Bilibili, etc.)

uv run analyze.py --video https://www.youtube.com/watch?v=xxxxx --prompt "总结视频"

uv run analyze.py --video https://www.youtube.com/watch?v=xxxxx --prompt "Summarize the video"

指定模型和抽帧数

Specify model and number of frames

uv run analyze.py --video video.mp4 --model doubao-vision --frames 20 --prompt "分析"

undefined

uv run analyze.py --video video.mp4 --model doubao-vision --frames 20 --prompt "Analyze"

undefined

注意事项

Notes

视频站点 URL 下载依赖
```
yt-dlp
```
，已作为 Python 依赖自动安装
抽帧模式下，帧数越多分析越详细，但 API 调用成本也越高
大视频文件下载可能需要较长时间，请耐心等待

Downloading video from website URLs depends on
```
yt-dlp
```
, which is automatically installed as a Python dependency
In frame extraction mode, more frames lead to more detailed analysis but higher API call costs
Downloading large video files may take a long time, please be patient