NoteWise is a Python CLI that converts YouTube videos, playlists, and batch files into local study materials. It reads metadata and transcripts, sends content through a LiteLLM-backed model, and writes files you control: Markdown notes by default, with optional HTML, PDF, DOCX, quizzes, and transcript exports.Documentation Index
Fetch the complete documentation index at: https://notewise.click/docs/llms.txt
Use this file to discover all available pages before exploring further.
Why NoteWise exists
Most video summarizers hide the result in a web app. NoteWise is built for people who want durable notes, scriptable workflows, and clear local state.Local study files
Notes are written to your output directory. Markdown is the default, and
rendered formats are optional.
Bring your own model
Use LiteLLM model strings for Gemini, OpenAI, Anthropic, Groq, OpenRouter,
Azure, Bedrock, ChatGPT OAuth, Copilot OAuth, and more.
Works for real YouTube inputs
Process watch URLs, Shorts, youtu.be links, playlists, bare IDs, and text
batch files.
Designed for operations
Inspect config, provider readiness, YouTube metadata, cache, logs, history,
and stats from the CLI.
What it produces
| Output | How to create it |
|---|---|
| Markdown notes | Default notewise process URL output. |
| HTML, PDF, DOCX | Use --format html, --format pdf, --format docx, or comma-separated formats. |
| Quiz file | Add --quiz. |
| Transcript export | Add --export-transcript txt or --export-transcript json. |
| Chapter folder | Add --chapter-directory-output when the video has YouTube chapters. |
Start here
Install NoteWise
Install the CLI with
uv, uvx, pipx, pip, Docker, or the short
notewise.click/install binary endpoint.Run the quickstart
Configure a provider, process one public video, and confirm the output file.
Process more than one video
Use playlists, batch files, concurrency, rate limits, and cache behavior
correctly.
Fix a failed run
Start with
doctor, info, config, cache checks, and recent logs.Source-backed defaults
These docs track NoteWise1.4.0. Source defaults are Python >=3.10, model gemini/gemini-2.5-flash, output directory ./output, output format md, temperature 0.7, max concurrent videos 5, YouTube request limit 10/min, chunk size 4000, overlap 200, and max concurrent chapters 3.