"Prompt engineering course" covers very different things. Some teach a marketer to get better drafts from a chat assistant. Others teach a developer to call a model API from code and test outputs. Picking the wrong one wastes weeks.
This guide helps you choose by starting from your job, not from a ranking. It gives a five-point checklist, shows what real course pages state and leave out, and lists questions to ask before you pay.
Why This Is Not a "Best Courses" List
We planned a ranked list of six courses. Before publishing it, we tried to confirm four facts for each course on its official page: fees, prerequisites, practical projects and who it suits.
We could not confirm all four for all six. Some pages list no price, or show prices only after sign-in or by region. A list with gaps would not help you decide, so this guide teaches you how to check those facts yourself.
Step 1: Start From What Your Job Needs
Prompt skills split into three broad tracks. Find yours first.
| Track | Typical roles | What you should be able to do after the course | Usually needs code? |
|---|---|---|---|
| Workplace user | Marketing, HR, operations, sales, analysts, managers | Get reliable drafts, summaries and analysis from chat assistants; check output; reuse prompt templates | No |
| Builder | Team leads, power users, product managers, analysts who automate | Design reusable prompts and simple assistants or workflows for a team; compare outputs systematically | Sometimes |
| Developer | Software, data and ML engineers | Call model APIs from code, structure outputs, use tools, and write prompt evaluations | Yes |
A workplace user does not need API lessons. A developer gains little from a course on using one chat app. Builders sit in between and should look for courses with templates, workflows and testing.
Step 2: Check Five Things on the Official Course Page
Use the course provider's own page, not a review site. Write down what you find, including what is missing.
1. Fees and What They Include
• Is the price stated, and is it one-time or a subscription?
• Does the certificate cost extra, or need a paid tier?
• Are there extra costs, such as API usage or tool subscriptions needed for exercises?
• Is financial aid or an employer plan available?
2. Prerequisites
• Does it require Python or another language?
• Does it assume an account with a specific AI tool?
• Is the level stated (beginner, intermediate, advanced)?
3. Practical Projects
• Does the page describe hands-on exercises or a final project?
• Will you build something you can reuse at work, such as a prompt library, an assistant or an evaluation script?
• Is the work graded or reviewed by anyone?
5. Currency
• When was the course last updated?
• Does it teach one vendor's tools or skills that transfer across models?
AI tools change quickly. A course built around an older interface may still teach sound principles, but check how much of it is tied to screens that may have changed.
What Real Course Pages State (and Leave Out)
To show the checklist in action, here is what four official pages stated when we read them on October 29, 2026. This is not a ranking or a recommendation, and details may have changed since.
| Course (provider page) | Fee stated on page | Prerequisites stated | Practical work stated | Who it suits, per page |
|---|---|---|---|---|
| Prompt Engineering Specialization, Vanderbilt University (Coursera) | No price shown; page says it cannot be taken for free and financial aid is available | Beginner level | "Multiple hands-on prompt engineering exercises" per course | Not stated as a role; beginner level; 3 courses, about 4 weeks at 10 hours a week |
| ChatGPT Prompt Engineering for Developers (DeepLearning.AI) | No price shown; an "accomplishment" requires a PRO plan | Basic Python | 9 lessons, 7 code examples in Jupyter notebooks using the OpenAI API | Beginner-friendly developers, also advanced ML engineers |
| Google AI Professional Certificate (Grow with Google) | $49 per month subscription in the US and Canada; may vary elsewhere | No experience required | 20+ hands-on activities; portfolio of AI projects | Anyone who wants to use AI at work; 8 courses of about an hour each |
| Anthropic courses (GitHub) | Materials are free to read; exercises use the Claude API, so API costs apply | Not stated as a level; starts with API fundamentals | Interactive prompt tutorial, real-world prompting, prompt evaluations and tool use | Not stated; content assumes working with the API |
Three lessons stand out:
• Price is often not on the page. Two of the four did not show one. Ask for it in writing before you enroll
• "Prompt engineering" in the title does not mean the same scope. The Google certificate covers prompting inside broader AI-at-work skills; the DeepLearning.AI course is a short, code-based developer course
• Free materials can still cost money. API-based exercises bill your account as you practice
Step 3: Match the Course to Your Track
Workplace users should look for: no coding, examples from business tasks, practice checking AI output and reusable templates. Short courses are fine if they include exercises.
Builders should look for: structured prompt patterns, ways to compare outputs, building simple assistants or workflows, and some exposure to evaluation.
Developers should look for: API use from code, structured outputs, tool or function calling, prompt evaluation and testing. Expect to pay some API costs.
If your role spans tracks, start with the lower one and add a second course later. Skipping foundations makes the advanced material harder.
Step 4: Ask These Questions Before You Pay
• What is the total cost, including certificate, subscription months and any API or tool fees?
• Can I see the syllabus and at least one exercise before enrolling?
• Is there a project I can adapt to my job, and does anyone review it?
• When was the content last updated, and for which models or tools?
• Is there a refund or cancellation policy?
• For teams: is there group pricing, progress reporting or a cohort option?
Red Flags
• Guaranteed job or salary claims. Prompt skills help, but no course can promise outcomes.
• No syllabus before payment. You should know what you are buying.
• Only videos, no exercises. Prompting is learned by doing and checking results.
• One-vendor tricks presented as universal rules. Good courses explain why a technique works.
• No update date. In a fast-moving field, undated content is a risk.
For Managers Choosing for a Team
Group your team by track before buying anything. A single course for everyone usually fits nobody well.
Pilot with three to five people per track. Ask each to apply one technique to a real task and report what changed. Use that, not completion rates, to decide whether to roll out.
Prompting courses build skills; they do not replace your company's AI usage rules. Make sure people know which tools and data are allowed before they practice on real work.
For broader options, see our guides to AI certification courses online and online AI bootcamps for working professionals.
Our Own Product Note
Disclosure: this is our own program. LuMay runs a live AI Masterclass covering building AI agents with tools including RAG, MCP and LangGraph. It is a builder and developer program, not a prompt-only course.
It is not ranked here. Check the page for current dates and fees.
Next Step
Write down your track, then check two or three course pages against the five-point checklist. Ask providers in writing for any fee or prerequisite the page does not state.
References
• Coursera, Prompt Engineering Specialization (Vanderbilt University): https://www.coursera.org/specializations/prompt-engineering
• DeepLearning.AI, ChatGPT Prompt Engineering for Developers: https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
• Grow with Google, Google AI Professional Certificate: https://grow.google/prompting-essentials/
• Anthropic, Educational courses (GitHub): https://github.com/anthropics/courses
