gemini-video-understanding

1# Gemini Video Understanding Skill
2 
3## Purpose
4This skill enables video understanding workflows using the Google Gemini API, including **video summarization**, **question answering**, **transcription with optional visual descriptions**, **timestamp-based queries (MM:SS)**, **scene/timeline detection**, **video clipping**, **custom FPS sampling**, **multi-video comparison**, and **YouTube URL analysis**.
5 
6## When to Use
7- Summarizing a video into key points or chapters
8- Answering questions about what happens at specific timestamps (MM:SS)
9- Producing a transcript (optionally with visual context) and speaker labels
10- Detecting scene changes or building a timeline of events
11- Analyzing long videos by clipping to relevant segments or reducing FPS
12- Comparing multiple videos (up to 10 videos on Gemini 2.5+)
13- Analyzing public YouTube videos directly via URL
14 
15## Required Libraries
16 
17The following Python libraries are required:
18 
19```python
20from google import genai
21from google.genai import types
22import os
23import time
24```
25 
26## Input Requirements
27- **File formats**: MP4, MPEG, MOV, AVI, FLV, MPG, WebM, WMV, 3GPP
28- **Size constraints**:
29  - Use inline bytes for small files (rule of thumb: <20MB).
30  - Use the File API upload flow for larger videos (most real videos).
31- **YouTube**:
32  - Video must be **public** (not private/unlisted) and not age-restricted.
33  - Provide a valid YouTube URL.
34- **Duration / context window** (model-dependent):
35…