100 Free AI Courses

Extensive live-recorded journey through generative models led by three AI experts.

Comprehensive generative AI course from foundational concepts to real-world deployment with hands-on projects.

Covers ML/AI foundations through advanced topics like recommender systems and predictive analytics, with a data science career roadmap.

Hands-on course on running LLMs locally with Ollama, building three projects including a RAG system and an AI recruiter.

Implements retrieval-augmented generation from scratch with LangChain, covering a dozen RAG papers with open-source notebooks.

Jeremy Howard's free code-first course teaching how to build and deploy deep learning models for vision, NLP, and tabular data with PyTorch and fastai.

Reproduces the GPT-2 124M model from an empty file, building the network then optimizing training speed, hyperparameters, and evaluation.

Builds neural networks from scratch in code, from backpropagation and micrograd up to a GPT and a tokenizer (includes the 'Let's build GPT' and 'Let's build the GPT Tokenizer' lectures).

Conference talk visualizing the transformer architecture and the attention mechanism that powers modern LLMs.

Visual explanation of how GPT-style transformers generate text, covering attention blocks and multilayer perceptron blocks.

Visual, intuition-first series on neural networks, gradient descent, and backpropagation, extended in 2024 with chapters on transformers and attention.

Covers LLM architecture, training techniques, LoRA, and Prefix Tuning with contributions from industry experts.

Free MLOps course on disciplined, collaborative, reproducible ML pipeline development.

Production-ready retrieval-augmented generation techniques: performance optimization, cost reduction, and accuracy improvement.

Vanderbilt's hugely popular non-coding course on crafting effective prompts and prompt patterns, free to audit (certificate paid).

The most popular course in the university's history with millions of learners: a no-math, no-code introduction to what AI is, what it can and cannot do, and its societal impact. Completely free.

Berkeley's core AI course on search, game playing, CSPs, probabilistic reasoning, MDPs, reinforcement learning, and machine learning. Lectures and Pacman projects free via YouTube and ai.berkeley.edu.

Free bootcamp on building and deploying LLM applications: prompt engineering, LLMOps, UX for language interfaces, and augmented language models.

Covers MDP planning, Q-learning, policy search, offline RL, exploration, and deep RL, with applications from robotics to healthcare. Full Spring 2024 lectures free on YouTube.

Stanford's flagship neural NLP course covering word vectors, transformers, post-training, and benchmarking. Spring 2024 lectures are free on YouTube.

Andrew Ng's canonical full-course lecture series covering supervised learning, SVMs, kernels, neural nets, and reinforcement learning. Current notes and problem sets live at cs229.stanford.edu.

A rigorous survey of current LLM research: architecture, data preparation, pre-training, post-training/alignment, and deployment. Reading list, slides, and materials openly available.

Free guides and handbooks on vector databases, embeddings, semantic search, and building RAG applications.

Official open-source collection of code examples and guides for the OpenAI API: prompting, function calling, embeddings, agents, and production patterns.

A 10-week, 20-lesson project-based curriculum covering data science principles, ethics, data preparation, visualization, and analysis.

A 12-week, 26-lesson curriculum on classic machine learning using Scikit-learn, with quizzes and project-based assignments.

21-lesson curriculum from Microsoft Cloud Advocates covering the fundamentals of building generative AI applications with hands-on code.

Meta-led course on prompting Llama 2 & 3, Code Llama, and Llama Guard, covering few-shot, chain-of-thought, and safe AI apps.

Efficient ML techniques: pruning, quantization, NAS, distillation, distributed training, and on-device LLM/diffusion deployment. Fall 2024 lectures free via efficientml.ai and YouTube.

LlamaIndex CEO-led course on building research agents with tool use, reasoning, and multi-document RAG.

LangChain Academy's flagship free course on LangGraph as an orchestration framework for complex agentic systems.

Practical tools for the responsible design of AI systems, covering bias, fairness, and model cards.

Interactive course building neural networks with TensorFlow and Keras, covering layers, activations, dropout, and batch normalization.

Hands-on micro-course covering how models work, validation, overfitting, and random forests, run entirely in Kaggle notebooks.

Community-built course covering computer vision fundamentals, CNNs, vision transformers, and multimodal models.

Study the theory and practice of diffusion models for image generation using the Hugging Face Diffusers library.

Learn to apply transformers to audio: speech recognition, audio classification, and text-to-speech generation.

Hands-on Deep Reinforcement Learning course covering Q-learning, policy gradients, PPO, and training agents in simulated environments.

Harvard's landmark intro to computer science and programming, covering algorithms, data structures, Python, SQL, and web development. Free to audit on edX.

Explores algorithms behind modern AI: search, knowledge representation, uncertainty, optimization, machine learning, and neural networks in Python. Free to audit on edX.

Google's practical AI-literacy course for using AI to boost productivity across any role, including prompting and responsible AI. Free to audit; financial aid available.

No-cost course covering what large language models are, their use cases, and methods like prompt tuning to improve their performance.

No-cost introductory course explaining what generative AI is, how it is used, and how it differs from traditional machine learning.

Google's refreshed practical introduction to ML with interactive lessons on regression, neural networks, embeddings, LLMs, and production ML.

Non-technical primer on what AI can and cannot do, how to spot opportunities, and how to build AI strategy in an organization. Free to audit.

Andrew Ng's non-technical overview of how generative AI works, how to use it day-to-day, and how to build an AI strategy for business. Free to audit.

Practical developer course on summarizing, inferring, transforming and expanding with LLM APIs, with runnable Jupyter examples. Free to access.

Build and deploy a serverless customer-support agent on Amazon Bedrock, examining its execution trace and reasoning steps.

Use Microsoft's AutoGen framework to build multi-agent conversational systems and agentic design patterns like reflection, tool use, and planning.

Automate complex business workflows by orchestrating a crew of role-playing AI agents that collaborate on tasks using the crewAI framework.

Build controllable agents from scratch and rebuild them in LangGraph, adding persistence, state management, and multi-thread conversation handling.

Cohere's free structured learning hub on embeddings, semantic search, RAG, prompt engineering, fine-tuning, and deployment.

CMU's graduate NLP course covering language modeling, word representations, RLHF, mixture-of-experts, interpretability, and complex reasoning. Full Fall 2024 lectures free on YouTube.

Anthropic's open-source course collection covering API fundamentals, prompt engineering, real-world prompting, tool use, and prompt evaluations.

A free, hands-on 9-chapter tutorial (Jupyter notebooks or Google Sheets) that teaches prompt engineering for Claude from basic structure to advanced complex prompts.

Part of the Gen AI 360 certification series, covering how to train, fine-tune, and deploy LLMs in production.

Free certification course on building production LLM apps with LangChain and vector databases.

Introduction to machine learning fundamentals through to generative AI concepts.

Roadmap course covering foundational math, key ML algorithms, LLMs, best practices, and AI career paths.

Beginner-friendly walkthrough of training a language model from start to finish.

freeCodeCamp course on integrating APIs, automating processes and building AI agents in n8n, including an email agent and research workflows.

Ongoing playlist covering LLM fine-tuning (LoRA/QLoRA), LangChain, Hugging Face, and end-to-end GenAI apps on AWS.

Walks through how transformer LLMs process text, from bag-of-words and Word2Vec to attention, with Hugging Face code examples.

Practical tour of the LLM ecosystem and real workflows across ChatGPT, Gemini, and Claude, including tool use, search, code interpreter, and multimodal interaction.

General-audience deep dive into the full LLM training stack: pretraining, tokenization, fine-tuning, hallucinations, and reinforcement learning.

Free no-code walkthrough of n8n covering core concepts, integrations and building AI agents, including an inbox-manager project with email categorization.

Course built with OpenAI on building production-grade AI agents combining reasoning models, tools, memory, and planning.

Berkeley RDI's open MOOC on LLM agents: reasoning and planning, math and coding, inference-time techniques, and agent safety and security, taught by leading researchers. Free livestreams, slides, and labs.

Builds a language model end-to-end: tokenization, transformer construction, GPUs/kernels, parallelism, scaling laws, inference, and evaluation. Spring 2025 lectures are free on YouTube.

A seminar series where leading researchers present the latest in transformer architectures and applications across LLMs, art, biology, and robotics. Lectures posted free on YouTube.

OpenAI's free, self-paced learning hub spanning AI fundamentals, applied AI, agents, and workflows, with certificates of completion.

Lesson-based course teaching the fundamentals of building AI agents, including agentic design patterns, tool use, and multi-agent systems.

Meta short course on building applications with the Llama 4 models and their multimodal and mixture-of-experts capabilities.

MIT's fast-paced bootcamp on deep learning fundamentals, CNNs, RL, and language models, with hands-on software labs. All 2025 lectures, slides, and labs are free online.

The original open, free prompt-engineering curriculum covering zero-shot, few-shot, chain-of-thought and more, used by millions of learners.

Quickstart teaching LangGraph's core components (State, Nodes, Edges, Memory) by building an email workflow.

Follow-up intensive from Google and Kaggle focused specifically on building, evaluating, and deploying AI agents.

Self-paced Learn Guide adapted from the record-breaking live course, covering foundation models, embeddings, agents, domain LLMs, and MLOps.

Beginner-to-expert course on building AI agents with smolagents, LlamaIndex, and LangGraph, with a completion certificate.

Free, ad-free course teaching LLMs and NLP using Transformers, Datasets, Tokenizers, and the Hugging Face Hub.

Built in partnership with Anthropic, teaches Model Context Protocol architecture and how to build and deploy MCP applications with a certification.

Beginner course on combining Google Cloud fundamentals with prompt engineering to dramatically improve Gemini responses.

Free Google-partnered course on building AI apps with the Gemini API and AI Studio across text, images, and code.

End-to-end course on designing, building, and deploying production multi-agent systems with the crewAI framework.

Andrew Ng's course on building agentic AI systems using the core design patterns of reflection, tool use, planning, and multi-agent collaboration.

Move beyond traditional OCR to agentic document extraction that understands layout, tables, and figures for structured data extraction.

Build an event-driven agentic workflow with LlamaIndex that parses documents, fills PDF forms, and retrieves data via RAG asynchronously.

Built with Anthropic, teaches the Model Context Protocol: how to build and deploy MCP servers that expose tools, resources, and prompts to AI apps like Claude Desktop.

Free project-based course on AI coding assistants, agents, MCP, n8n automation and CI/CD, with a certificate on completion.

Continuously updated open reference on prompting, context engineering, RAG and agents with copyable examples and model-specific guidance across many languages.

Applies AI tools without coding to automate workflows and analyze data for business, part of the No-Code AI for Business specialization. Free to audit.

Getting-started course on the Model Context Protocol, plus an advanced-topics follow-up, for connecting Claude to tools and data.

Learn to build, configure, and share Skills in Claude Code: reusable instructions Claude applies to the right tasks automatically.

Hands-on course on getting up to speed with Claude Code, Anthropic's agentic command-line coding tool.

In-depth Anthropic Academy course on building applications with the Claude SDK and Anthropic API.

Comprehensive project-driven course on building and deploying agentic AI with LangGraph, including human-in-the-loop, RAG, and streaming.

Fundamentals of retrieval-augmented generation and the Model Context Protocol, including building your own MCP server and client in Python.

End-to-end production RAG pipeline covering vector DB optimization, observability, and agentic and multimodal architectures.

Free course on using AI for marketing strategy, content creation, personalization and analytics, with strategic prompting and responsible-use guidance.

Google's five-step prompting course for everyday work (email, data analysis, presentations), free via Google Skills.