Getting Started with Shirube YouTube Transcript Tool
Go from downloading the portable ZIP archive to inspecting your first 10 extracted and translated video transcripts in under five minutes.
Download & Extract the Portable ZIP
Download Shirube-YouTube-Transcript-Tool-v1.1.1-win64.zip and extract it to a permanent folder such as C:\Tools\Shirube-YouTube-Transcript-Tool-v1.1.1. Do not run the script directly from inside an unextracted ZIP preview window.
Launch via Start-App.bat
Double-click Start-App.bat. On your very first launch, the launcher verifies that Python 3.10+ is installed and automatically installs the 5 lightweight packages listed in requirements.txt (customtkinter, yt-dlp, youtube-transcript-api, requests, curl_cffi).
Always Run “🧪 Test First 10” Before a Full Channel Crawl
Paste your target channel URL (for example, https://www.youtube.com/@TED), choose an Output Folder, optionally select a 🌐 Translate To language (e.g., English (en) or Japanese (ja)), and click 🧪 Test First 10.
Why start with Test First 10? It indexes only the first batch of videos and extracts 10 transcripts in ~20 seconds so you can verify caption availability and CSV formatting before running a multi-hour crawl on a 1,000-video channel.
Click “▶️ Extract All / Resume” & Open Output Folder
Once you are happy with the 10-video sample, click ▶️ Extract All / Resume. The tool automatically skips the 10 videos already saved in channel_cache.db and processes the remaining videos on the channel. Click 📂 Open Folder at any time to inspect videos.csv, transcripts.csv, and the transcripts/ folder.