Quick Start
Install NoteWise and generate your first study notes in under 3 minutes.
Installation
All installation methods: uv, pipx, pip, standalone binary, and Docker.
CLI Reference
Every command, subcommand, and flag — with defaults and real examples.
Configuration
Configure your API key, model, output directory, concurrency, and more.
What NoteWise does
Fetch the transcript
NoteWise retrieves the YouTube transcript (with chapter markers if available) using yt-dlp — no browser required.
Chunk and process
For long videos, the transcript is split into overlapping chunks. Chapter-aware videos get dedicated notes per chapter.
Generate notes with your LLM
Each chunk is sent to the configured LLM (Gemini, OpenAI, Anthropic, Groq, and more) to produce structured Markdown — headers, definitions, examples, and key concepts.
Key features
Multi-provider LLM support
Works with Gemini, OpenAI, Anthropic, Groq, Mistral, Cohere, DeepSeek, and xAI via LiteLLM. Swap providers with a single flag.
Chapter-aware processing
Automatically detects video chapters and generates separate, structured notes per chapter for long lectures and courses.
Batch processing
Process a full playlist or a
.txt file of URLs in one command, with configurable concurrency.Quiz generation
Optionally generate a ready-to-use multiple-choice quiz alongside each set of study notes.
Local cache
Transcripts and run stats are cached in SQLite. Already-processed videos are skipped automatically.
CI/cron friendly
A
--no-ui flag disables the Rich dashboard and outputs plain progress lines for scripting and automation.Supported providers
NoteWise routes all LLM calls through LiteLLM, so any model string LiteLLM supports works out of the box. The default model is Gemini 2.5 Flash — get a free API key at aistudio.google.com.LLM providers
See all supported providers and how to configure each one.
Configuration schema
Full reference for every configuration key and its default value.