skills.rest
levanminhduc
levanminhducCOMMUNITY
Explore all skills by this author →
GitHub

gemini-video-understanding

Analyze local and YouTube videos with the Gemini API for summarization and Q&A.

Updated Jul 23, 2025
Search Tags:#summarization#qa#youtube#transcription#gemini#video-analysis
One-click install
npx skills add https://github.com/levanminhduc/LuongHoaThoNew --skill gemini-video-understanding-levanminhduc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gemini-video-understanding
Source: https://github.com/levanminhduc/LuongHoaThoNew/tree/main/.claude/skills/gemini-video-understanding
Command: npx skills add https://github.com/levanminhduc/LuongHoaThoNew --skill gemini-video-understanding-levanminhduc

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires google-genai, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill enables automated, AI-powered analysis of video content using Google's Gemini, including summarization, Q&A, transcription with visual context and timestamps, and flexible input methods (local files, YouTube), with support for multiple models and long context windows.

Core Features & Use Cases

  • Video Summarization: Create concise summaries of video content
  • Transcription with Visuals: Transcribe audio with timestamps and on-screen descriptions
  • Question Answering: Answer questions about video content
  • Video Clipping: Analyze specific segments with start/end offsets
  • Multi-Video Comparison: Analyze and compare multiple videos
  • YouTube Support: Process YouTube videos directly
  • Flexible Input Methods: Local files, inline data, File API, YouTube
  • Model Choice & Frame Rate: Select Gemini models and custom FPS

Quick Start

Run the sample: python .claude/skills/gemini-video-understanding/scripts/analyze_video.py --video-path path/to/video.mp4 --prompt "Summarize this video in 3 points"

Frequently Asked Questions about gemini-video-understanding

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze and summarize video content using AI?▼

Video summarization with AI uses computer vision to extract key points from footage. This Skill applies Gemini's vision capabilities to local files and YouTube videos, generating concise summaries, transcriptions with timestamps, and visual descriptions in a single analysis.

Can I transcribe videos with visual context and timestamps?▼

Yes. Transcription with visuals captures both audio and on-screen descriptions aligned to timestamps. This Skill processes local video files and YouTube URLs through Gemini, delivering frame-aware transcripts that describe what viewers see alongside dialogue.

Does this support YouTube videos or only local files?▼

Both. This Skill accepts local video files in nine common formats, YouTube URLs, and multiple videos up to 10 files per run. It resolves API credentials from environment variables or .env files and handles format conversion automatically.

How do I answer specific questions about video content?▼

Question answering analyzes video segments to extract answers from visual and audio information. This Skill lets you query specific timestamps or entire videos using Gemini's long-context window, returning answers grounded in the footage analyzed.

What video formats and models does this Skill support?▼

This Skill handles nine common video formats with configurable frame rates and supports Gemini 2.5 and 2.0 series models. You select the model and set custom FPS, start/end offsets, and processing parameters via command-line arguments.

Can I analyze specific segments of a video with custom start and end points?▼

Yes. Video clipping with offsets lets you define start and end times for focused analysis. This Skill processes those segments independently, applying summarization, Q&A, and transcription only to the portion you specify.

Related Skills

Explore More Data & Analytics Skills→
benjaminshafiibenjaminshafii
10

bitwarden

Configure Bitwarden CLI with API key authentication to retrieve vault credentials.

Community
Intermediate
freestyleflyfreestylefly
2

note-weaver-skill

Convert fragmented multi-source notes into structured Markdown and JSON knowledge graphs.

Community
Advanced
ahostbrahostbr

ls-youtube-transcript

Extract YouTube subtitles into timestamped markdown transcripts via yt-dlp.

Community
Intermediate
dingtom336-gifdingtom336-gif
22

viral-video-replicator

Reverse-engineer videos into Seedance 2.0 prompts with FFmpeg and ASR.

Community
Advanced