Learn Prompting

Learn Prompting is an online school for generative AI, built around prompt engineering and AI security. It serves individual learners with free courses and a paid Plus subscription, and companies with Teams and Enterprise plans. Its founder and CEO is Sander Schulhoff, who also runs HackAPrompt; Fady Yanni is co-founder and COO. The home page says the platform is trusted by "the world's biggest companies and 3,000,000+ people".

Where it came from

The About page tells the origin story. Learn Prompting began as a college English assignment: Schulhoff read a few hundred papers on prompting over 2 weeks, built a site and released it in October 2022, before ChatGPT launched. The page says the guide quickly became an essential resource in the field, and that it is cited by Google, Wikipedia, O'Reilly and Scale AI and used internally to train employees at companies such as Google, Microsoft, OpenAI and Meta. That original guide, published as the Docs, is still free and open-source. The company estimates that about 50% of other prompt engineering courses and guides draw on it, and credits open-source contributions from 100+ users. The paid courses came later, which the About page explains plainly as a way "to keep the lights on" while offering more in-depth video content and certification.

What a learner gets

The course catalog is sorted into beginner, intermediate and advanced levels. The titles listed across the site are ChatGPT for Everyone, Introduction to Prompt Engineering, Advanced Prompt Engineering, Introduction to Generative AI, Introduction to Large Language Models (LLMs), Introduction to Generative AI Agents, Introduction to Retrieval-Augmented Generation (RAG), AI Safety, Runway ML for Everyone, Introduction to Generative AI in Marketing, Boost Your Day-to-Day Efficiency With Generative AI, Introduction to Prompt Hacking and Advanced Prompt Hacking. That spread runs from someone learning to use ChatGPT at work, through marketers and image and video makers, to developers working on agents and RAG.

The newest format is the Specialization. A banner across every page announces the Google AI Studio Specialization, in which learners build 12 real apps with no code needed. The menu describes Specializations as a way to build real apps with AI tools.

The flagship advanced course is the AI Red-Teaming and AI Security Masterclass, offered on demand. The home page says it is led by experts in AI and cybersecurity, features Sander Schulhoff and guest speakers, and teaches how to attack AI systems to make models safer. Schulhoff's background gives this course its weight: the About page says he partnered with OpenAI to run the first AI red-teaming competition, HackAPrompt, and that HackAPrompt now produces security research with frontier AI labs.

Inside the paid courses, Plus members get an in-course AI Playground with an AI Tutor, so prompts can be tried against a model within the lesson, plus hand-graded AI projects. Members also join what the pricing page calls "the largest AI Learning Discord community"; the About page counts 40K+ Discord members.

Free learning

Much of the school costs nothing. The open-source Docs remain free, the Free option on the pricing page gives access to the free courses with certificates of completion, and there is a free email course, Learn Prompt Hacking in 7 Days, covering why LLM security matters (day 1), prompt hacking types and intents (days 2 and 3), common attacks and defense strategies (days 4 and 5), and future challenges (days 6 and 7). The resources menu also lists a Prompt Hacking Guide, a vocabulary page and the help center.

Certificates and certifications

The site separates two things. Every course earns a certificate of completion, on the Free option for free courses and on Plus for the whole library, and the help center has an article on the difference between certificates of completion and certificates of proficiency. Separately, the certifications page offers three exams it calls industry-recognized, all at professional level: the AI Red Teaming Professional Certification (AIRTP+), described as mastering advanced techniques in AI security assessment; the Certified Prompt Engineering Professional, on creating effective, secure and optimized prompts; and the Certified AI Engineering Professional, on building, deploying and optimizing AI systems. These are the company's own credentials, and the page links to team certification through Learn Prompting for Teams.

Prices and plans

The pricing page lists four plans, with a monthly and annual toggle and a "Save 49%" label on annual billing:

• Free, for individuals getting started: free courses and certificates of completion.
• Plus, $21 a month billed annually: the entire course library including Plus courses, certificates of completion for all courses, the AI Playground with AI Tutor, hand-graded projects, the Discord community and priority support.
• Teams, for 2 to 99 users at $25 per seat per month billed annually: the Plus features for each seat, plus basic administration.
• Enterprise, for 100+ users, priced through sales: personalized learning paths per department with tailored projects, an admin dashboard, a dedicated customer success manager, advanced analytics and reporting, single sign-on through Okta, Auth0, Azure and others, and LMS/LXP integrations.

The pricing page prints the annual-billing rates. The help center's Subscriptions and Payments section has articles on payment methods, cancellation and the refund policy for Plus.

Research behind the courses

The home page sells the courses as "research-backed" and says its instructors are leading generative AI experts, professionals and researchers. The About page backs this with The Prompt Report, which Schulhoff led: a 76-page survey of prompt engineering co-authored with OpenAI, Microsoft, Google, Princeton, Stanford and other institutions, analyzing 1,500+ academic papers and covering 200+ prompting techniques. It also says the team's research on prompting and LLM security has been cited by OpenAI in a paper on mitigating prompt injections and by NIST in its adversarial AI paper, and that the free guide is updated with this material.

For companies

For a training team, Learn Prompting works as a focused AI library. Teams suits a department of up to 99 people who need seats, the full catalog and simple administration at a stated per-seat price. Enterprise adds the pieces an L&D function expects: department-level learning paths, reporting, SSO and connection to an existing LMS or LXP. The home page pitches this as a way to teach employees the latest techniques from its research.

Who it suits

Learn Prompting fits anyone who wants to get properly good at working with language models without first learning to code: office workers and marketers starting from ChatGPT for Everyone, builders using the no-code Google AI Studio Specialization, developers moving into agents and RAG, and security people drawn to the red-teaming masterclass and the AIRTP+ exam. The free guide and free courses let a learner test the teaching style before paying $21 a month on annual billing. The caveat is scope: this is a generative AI school, so a learner after classical statistics or a general data science curriculum will find only the LLM side of the field here.