Two different kinds of course
Generative AI courses split into two groups that are easy to confuse. Usage courses teach what the technology can and cannot do, how to write prompts, and how to plan a project; they need no programming. Building courses teach you to call a model from code, handle responses, add retrieval or tools, and evaluate results; they assume you can already program.
Generative AI for Everyone is a clear example of the first group: the course page says it requires no prior knowledge of AI or coding, lists about 6 hours, and includes an optional 60-minute module on trying generative AI code [1]. Generative AI with Large Language Models is in the second group: it is listed as intermediate, says you should have some Python experience and basic machine learning concepts, and includes three labs of 120 minutes each on summarization and fine-tuning [2].
Options compared
We have not completed these courses. Details are from the official pages checked on 30 October 2026.
| Option | Type | Prerequisites stated | Projects or labs | Cost to check |
|---|---|---|---|---|
| Generative AI for Everyone (Coursera, DeepLearning.AI) | Usage | No AI or coding knowledge needed [1] | Exercises; optional code module [1] | Check the course price, or whether it is included in Coursera Plus at $59 a month or $399 a year [6] |
| Generative AI with Large Language Models (Coursera, DeepLearning.AI and AWS) | Building, model-focused | Some Python and basic machine learning [2] | Three 120-minute labs including fine-tuning [2] | Coursera pricing; Plus is $59 a month or $399 a year [6] |
| Become an LLM Engineer (Educative skill path) | Building, application-focused | Not stated on the page [3] | Practical work with OpenAI APIs, vector databases, LangChain, RAG, LoRA and QLoRA, CrewAI; 15 hours, 93 lessons [3] | Educative plan; price not shown on page [3] |
| Claude quickstart (Anthropic docs) | Building, first API call | Console account and API key [4] | First API call in cURL, CLI, Python, TypeScript, C#, Go, Java, PHP or Ruby [4] | Docs free to read; check the provider API pricing before building |
| OpenAI developer quickstart | Building, first API call | API key from the dashboard [5] | First request with SDKs for JavaScript, Python, .NET, Java, Go, Ruby [5] | Docs free to read; page says to add credits to keep building [5] |
Start with vendor documentation if you already code
If you can write a small Python or JavaScript script, the vendor quickstarts are the fastest way to find out whether you like this work. Anthropic's quickstart lists two prerequisites, a Console account and an API key, and walks through a first call to the Messages API in several languages [4]. OpenAI's quickstart starts with creating an API key, exporting it as an environment variable, installing an SDK and running a first request; it notes you add credits to keep building [5].
An evening with a quickstart tells you whether you need a course at all. If you get stuck on environment variables or package installs, the gap is general programming, not AI, and a beginner programming course is the better next step.
What to look for in a building course
- Stated prerequisites you actually meet, for example Python experience for the Coursera LLM course [2].
- At least one project that calls a real model API and handles its response, not only slides.
- Coverage of retrieval (RAG) or tool use if you plan to build applications; the Educative path lists both RAG and agents [3].
- Evaluation: how to tell whether the output is good enough, not only how to get output.
- A clear statement of whether API usage costs are on you.
Skip list
- You do not write code and do not plan to: skip building courses and vendor quickstarts; a usage course such as Generative AI for Everyone fits [1].
- You have never used Python: skip the intermediate LLM course until you have a beginner Python course behind you; its FAQ points no-programming learners to a beginner course first [2].
- You only want to build one small feature at work: try the quickstart first; a 15-hour path may be more than you need [3] [4].
- You want to train models from scratch: these courses focus on using and adapting existing models, not research-level training.
Worked example: two learners (invented)
A product manager wants to understand what generative AI can do for their team. They take the 6-hour usage course over three weeks at 1 to 2 hours a week, as the course page suggests [1], and stop there.
A backend developer who writes Python spends one evening on a vendor quickstart [4] [5], then wants structure for retrieval and agents. They compare the Educative path, which lists RAG and agent frameworks [3], with the Coursera LLM course, which focuses on fine-tuning labs [2], and pick based on whether they want application building or model adaptation.
Questions before you decide
Do I need to know Python to learn generative AI?
Not for usage courses: Generative AI for Everyone says it needs no AI or coding knowledge. For building courses, yes; the Coursera Generative AI with Large Language Models page says you should have some Python experience.
Are vendor API docs enough to start building with generative AI?
They are enough for a first working call. The Claude and OpenAI quickstarts walk through getting an API key, installing an SDK and making a request. A course helps when you want structured projects on retrieval, agents or evaluation.
Sources & verification
Product details and prices can change. Check the linked provider before buying.
- Coursera: Generative AI for Everyone (DeepLearning.AI) Accessed 30 Oct 2026
- Coursera: Generative AI with Large Language Models (DeepLearning.AI and AWS) Accessed 30 Oct 2026
- Educative: Become an LLM Engineer skill path Accessed 30 Oct 2026
- Claude Developer Platform: Get started with Claude Accessed 30 Oct 2026
- OpenAI API: Developer quickstart Accessed 30 Oct 2026
- Coursera Plus: plan page Accessed 30 Oct 2026
Sources link directly to providers. Product buttons may use separately labeled affiliate links. Read our disclosure.