Learn AI: 25 courses and guides from vendors and universities

Before you pick one

1. Pick one goal. Understanding AI, using AI tools well, and building with AI are different paths. Start in the matching section below.
2. Check what "free" covers. Lessons may be free while graded work, certificates or full access cost money.
3. Check the prerequisites. The build and deeper sections assume some Python.
4. Check the language. We mention a language only where the provider's page showed it. If a row says nothing, look at the course page.
5. Remember whose view it is. A course from an AI company mostly teaches that company's own tools, so it helps to mix in a university or independent source.

Start here: understand AI, no coding5

Plain-language courses on what AI is, what it can and cannot do, and how to work with it responsibly.

• Elements of AIUniversity of Helsinki / MinnaLearn · CourseA non-technical introduction to what AI is, what it can and cannot do, and how it shows up in daily life, followed by a second, more technical course called Building AI.Cost: Free. The site calls both courses free online courses.Self-paced. The page says Introduction to AI needs no complicated math or programming, and recommends some basic Python for Building AI. A Spanish page exists for the introduction; the site says Building AI is in English.
• AI Fluency: Framework & FoundationsAnthropic Academy · CourseTeaches a framework for working with AI in ways that are effective, efficient, ethical and safe. Developed with Prof. Joseph Feller and Prof. Rick Dakan.Cost: Free. The register button reads FREE.Online lessons with downloadable PDF materials. The page says it suits newcomers and experienced practitioners. A certificate of completion is offered after a final assessment.
• AI Capabilities and LimitationsAnthropic Academy · CourseAn introductory course on how generative AI behaves and why, covering next-token prediction, knowledge, working memory and steerability.Cost: Free. The register button reads FREE.Online lessons with try-it-out activities and a course quiz. The page calls it introductory and says it complements the AI Fluency framework course.
• AI Fluency for studentsAnthropic Academy · CourseTeaches students to use AI responsibly for learning, career planning and academic success, using the AI Fluency framework.Cost: Free. The register button reads FREE.Organized in modules. The page says learners should understand the AI Fluency framework, which Anthropic's core AI Fluency course covers. A certificate of completion is offered after a final assessment.
• AI Fluency (Microsoft Learn path)Microsoft Learn · Learning pathA beginner path covering AI basics, generative AI, search and prompting, responsible AI, and getting started with Microsoft Copilot.Cost: The page does not state a cost.7 modules, level Beginner. The path is centered on Microsoft Copilot. Module badges and a path trophy are shown; the page does not mention a certificate. The page metadata says it was last updated 2025-08-12.

Use AI tools well6

Courses and short guides on getting better results from chat assistants, written by the companies behind them.

• OpenAI Academy coursesOpenAI · Course hubSelf-paced courses in four tracks: apply AI at work, build with AI, lead AI adoption, and teach and learn with AI. The foundations courses cover ChatGPT and prompting basics.Cost: Free, according to an OpenAI help article that counts 14 free courses. A ChatGPT account is needed to enroll and save progress.Estimated 30 to 180 minutes per course. A badge goes to those who score 80% or higher on the assessment; OpenAI says neither the badge nor the pathway certificate is a certification.
• Claude 101Anthropic Academy · CourseIntroduces how to use Claude for everyday work tasks, with its core features and pointers to more advanced topics.Cost: Free. The register button reads FREE.Online lessons with progress tracking and quiz scores; the curriculum lists about 14 lessons in 5 modules (our count). A certificate of completion is offered.
• Prompting guide 101Google (Gemini for Workspace) · GuideA quick-start web guide to writing prompts for Gemini: persona, task, context and format, six short tips, and example prompts by job role.Cost: Free to read. The page needs no form.The examples are work roles rather than school subjects.
• Prompt engineering best practices for ChatGPTOpenAI Help Center · GuideA plain-language help article on writing prompts for ChatGPT: be clear and specific, refine over several tries, and describe the tone you want.Cost: Free to read.
• Copilot PromptsMicrosoft · GuideA public page of example prompts for common work tasks in Word, Excel, PowerPoint and Outlook, with basics on what a prompt is and how to improve results.Cost: The page can be viewed without signing in. Trying a prompt opens Copilot Chat, which probably needs a sign-in (the page does not say).The page says it is available in many languages, including Spanish.
• Prompting best practicesAnthropic (Claude Platform docs) · GuideA reference for writing prompts for Claude: being clear, adding context, giving examples, structuring prompts with XML tags, assigning a role, plus tips for agent-style work.Cost: Free to read, no account needed.Written for developers and updated often, so details can change.

Build with AI and agents (some coding)6

Hands-on courses for making apps and agents. Most assume basic Python.

• Generative AI for BeginnersMicrosoft (GitHub) · Course on GitHubA 21-lesson course on the fundamentals of building generative AI apps, from large language model and prompting concepts to hands-on code projects in Python and TypeScript.Cost: Free to read on GitHub. Running the code needs a GitHub account and a model provider; the page does not state costs.The page says it is translated into 50+ languages, including Spanish. It also says GitHub Models retires at the end of July 2026 and recommends Microsoft Foundry Models, so setup steps may differ.
• AI Agents for BeginnersMicrosoft (GitHub) · Course on GitHubA beginner course on building AI agents in 18 standalone lessons that combine a written guide, a short video and Python code.Cost: Free to read. The code examples need a Microsoft Foundry / Azure account, and the page does not state what that costs.We opened a translated copy of the course page (Swedish), which lists 54 translations including Spanish and says the English original is the reference. It suggests starting with Generative AI for Beginners if you are new to generative AI.
• AI Agents CourseHugging Face · CourseExplains how AI agents work (tools, language models, chat templates) and how to build them with frameworks such as smolagents, LangGraph and LlamaIndex.Cost: Free. The page says the certification process is completely free too.Needs basic Python, basic knowledge of language models and a free Hugging Face account. About 3 to 4 hours a week per chapter. Two free certificates. The language selector lists Spanish; we did not check how complete the Spanish version is.
• LLM CourseHugging Face · CourseA technical course on large language models and NLP with Hugging Face libraries, from using and fine-tuning models to building demos.Cost: Free. The page says it is free and without ads.Requires good Python knowledge. Each chapter is designed for a week at about 6 to 8 hours. The page says there is no certification. Spanish is listed among the translations marked work in progress.
• Introduction to LangChain: Build AI Agents with PythonLangChain Academy · CourseAn introduction to building AI agents with the LangChain framework, from a first agent to debugging with LangSmith.Cost: Free. The academy home page says its course content is always free.30 lessons in 3 modules, about 1.5 hours of video, with hands-on projects. Aimed at people who code.
• Anthropic AcademyAnthropic · Course hubThe home page of Anthropic's courses. When we looked it listed 23 courses, including Building with the Claude API and Introduction to Model Context Protocol (both free and needing Python, according to their pages) and a course on agent skills.Cost: The courses we opened showed a free register button; check each course page.

Go deeper: university and independent courses3

Longer courses on how machine learning works. They ask for more time and more programming.

• CS50's Introduction to Artificial Intelligence with PythonHarvard University (CS50) · CourseA Python-based course on the ideas behind modern AI: search, knowledge, uncertainty, optimization, machine learning, neural networks and language, taught through projects.Cost: Free to take through the CS50 site. Optional paid certificates exist; Harvard Online's page lists $299 for a verified certificate.7 weeks of material. The prerequisite is CS50x or at least one year of Python, so it is not an absolute-beginner course. Harvard Online lists 10 to 30 hours a week.
• Practical Deep Learning for Codersfast.ai · CourseA coding-first course on applying deep learning to practical problems such as computer vision, language, tabular data and recommendations, using PyTorch and fastai.Cost: Free. The page calls it a free course for people with some coding experience.9 lessons of about 90 minutes each, plus a free book. The page asks that you know how to code and have finished a high school math course. It does not mention a certificate.
• 6.034 Artificial Intelligence (Fall 2010)MIT OpenCourseWare · University lecturesRecorded undergraduate lectures introducing knowledge representation, problem solving and learning methods in AI, taught by Prof. Winston.Cost: Free. The site describes OpenCourseWare as freely shared knowledge.The course is from 2010, so it predates today's large language models and covers classic AI. No certificate or credit is mentioned. Content is licensed CC BY-NC-SA 4.0.

Free to start, paid for more5

These start free or at no listed charge, but a certificate, graded work or full access costs money, or the page does not make clear what is free.

• Google AI EssentialsGoogle (on Coursera) · CourseA beginner program on generative AI basics, writing prompts, using AI tools for everyday work with Gemini, and using AI responsibly.Cost: A 7-day free trial, then a Coursera subscription of $49 a month in the US and Canada; the price varies by country and financial aid is available.No experience required. Under 5 hours across 5 modules on Google's page; Coursera shows different totals. A shareable Google certificate is offered. Google's pages differ on languages, so check before enrolling.
• AI for EveryoneDeepLearning.AI · CourseA non-technical course on basic AI terms, how organizations can apply AI, and AI's effect on society, taught by Andrew Ng.Cost: Paid. The page says $49 for 180 days of certificate eligibility and mentions no audit option; we could not open the Coursera version.Level Beginner, no technical or business background required. 35 video lessons, 6h54m of content.
• AI Prompting for EveryoneDeepLearning.AI · CourseA beginner course on using AI chat tools to find information, brainstorm, write, and create with multimedia and code.Cost: The page shows no standalone price. The certificate and graded assignments are tied to PRO membership, which the FAQ lists at $25 a month billed annually or $30 a month billed monthly.Level Beginner. 21 video lessons, 7h4m, in 3 modules. We could not tell whether the videos can be watched without PRO.
• Introduction to Generative AIGoogle Skills (Google Cloud) · CourseA short introduction to what generative AI is, how it differs from traditional machine learning, and Google tools for building generative AI apps.Cost: The course page does not list a cost. Google's AI hub lists it at no charge, and describes the wider Google Skills catalog as 35 monthly learning credits at no cost, with full access at $29 a month.Introductory. 45 minutes on the course page (Google's hub says 30). A badge is offered.
• New to AI (AWS Skill Builder)Amazon Web Services · Course hubAn AWS page for people new to AI that links short Skill Builder courses on machine learning and generative AI, a prompt engineering course and the no-code PartyRock app builder.Cost: Skill Builder has a free tier, and AWS says it offers 1,000+ learning resources at no cost, with paid subscriptions ($29 a month) for extras. The page does not say which courses are free, so check each one.