-
Notifications
You must be signed in to change notification settings - Fork 4
FAQ
Jaydeep Solanki edited this page Mar 8, 2026
·
8 revisions
Quick answers for common usage, config, and output questions.
It is a Python CLI that converts YouTube videos, playlists, and URL batches into Markdown study notes using transcripts plus LLM-based structuring.
- one Markdown file per video for standard runs
- one Markdown file per chapter for long videos with chapters
- nested output when playlists are processed
The CLI is open source. Model/provider usage may still cost money depending on your API plan.
- single YouTube video URLs
- playlist URLs
- text files with one URL per line
yt-study config-pathyt-study setup --force~/.yt-study/config.env
No. Runtime config comes from ~/.yt-study/config.env plus environment variable overrides.
DEFAULT_MODELOUTPUT_DIRMAX_CONCURRENT_VIDEOSTEMPERATUREMAX_TOKENSGEMINI_API_KEYOPENAI_API_KEYANTHROPIC_API_KEYGROQ_API_KEYXAI_API_KEYMISTRAL_API_KEY
| Model family | Key |
|---|---|
| Gemini / Vertex | GEMINI_API_KEY |
| GPT / OpenAI | OPENAI_API_KEY |
| Claude / Anthropic | ANTHROPIC_API_KEY |
| Groq | GROQ_API_KEY |
| Grok / xAI | XAI_API_KEY |
| Mistral | MISTRAL_API_KEY |
When both are true:
- the video is longer than
3600seconds - chapters are available
The pipeline falls back to the raw video ID for naming.
The pipeline continues with the remaining videos and reports failures in the final summary.
Default:
~/.yt-study/logs/
Fallback:
./logs/
Use Windows Terminal or PowerShell 7, or switch to UTF-8:
chcp 65001- Usage for command examples
- Configuration for config details
- Troubleshooting for failure recovery
- Development for contributor workflow