Getting Started

For the cross-course onboarding (registration, account setup, calendar, newsletter), read Joining a Cohort first.

This page covers the LLM Zoomcamp specifics.

Star the GitHub repository

github.com/DataTalksClub/llm-zoomcamp

Star it so you can find it later. All course materials are here, with each module having its own folder. Cohort-specific homework, deadlines, and the launch-stream link are under cohorts/2026/.

If a lot of new participants star the repo around launch time, the repo can surface on GitHub Trending and pull more people into the cohort. It is the easiest way to support the course.

Join the LLM Zoomcamp Slack channel

After joining the DataTalks.Club Slack workspace, find the channel:

#course-llm-zoomcamp

This is your primary support and Q+A channel for the course. Use threads when replying. Don’t tag instructors directly - other participants often answer first, and tagging discourages that.

t.me/llm_zoomcamp is the announcement-only channel. Telegram is the most reliable place to catch important updates because Slack gets noisy. Telegram announcements are auto-reposted to Slack, so you do not strictly need it.

Subscribe to YouTube

The lectures are pre-recorded and live in two playlists on the DataTalks.Club YouTube channel:

  • LLM Zoomcamp main playlist: the canonical pre-recorded module videos, reused across cohorts.
  • LLM Zoomcamp 2026: cohort-specific recordings (launch stream, pre-course Q+A, office hours).

For all the playlist links, see Resources.

Bookmark the FAQ

The LLM Zoomcamp FAQ collects answers to module-specific and technical questions from previous cohorts. Check it before posting in Slack.

There is also a community-built bot that answers questions using the FAQ and Slack history; the link is pinned in the course Slack channel.

Pick your LLM provider

The course uses OpenAI in its examples. You are free to use any provider for your own work and project. Practical options:

  • OpenAI: the default. A few dollars of credit covers the whole course.
  • Groq: free tier is generous enough to complete the course end to end.
  • Ollama or similar: for fully local, no-API workflows on a normal laptop.

For tradeoffs and the recommended setup, see Environment Setup.

Set up your environment

The course uses Python with Docker for the supporting services (search engines, monitoring), and notebooks for most of the work. You will install:

  • Python (3.10 or newer).
  • Docker.
  • Jupyter (covered in module 1).
  • An LLM provider account (OpenAI key, Groq key, or both).

For the choices you need to make, see Environment Setup.

Optional: pre-course search engine workshop

The course’s RAG pipeline is built on top of search. Last cohort recorded a Build Your Own Search Engine workshop that walks through how a small in-memory search engine works. It is not required for module 1, but the implementation it produces is reused later in the course. Worth one or two evenings if you have time before the cohort starts.

Start Module 1

Begin with Module 1: Introduction to LLMs and RAG. Module 1 covers what LLMs are, what RAG is, and walks you through your first end-to-end RAG implementation. For what each module covers, see Curriculum.

Where to look next