We earn a commission if you buy through links on this page. How we're funded.
Quick answer
Generative AI with Large Language Models, the DeepLearning.AI and AWS course on Coursera, is a compact, technical tour of how language models are pre-trained, fine-tuned, aligned and put into applications, about 17 hours by Coursera's module cards, and we rate it 3.7 out of 5. It suits developers and data scientists with some Python and machine-learning basics who want to understand the model side of generative AI. It is not for non-coders, and not for anyone who wants to build agents or retrieval systems: Coursera dates the course to 2023 and states no update since, and its three labs are guided rather than a project.
Why we score it 3.7 / 5
A DeepLearning.AI and AWS course on how large language models are trained, adapted and deployed: transformers and scaling laws, instruction fine-tuning and LoRA, reinforcement learning from human feedback, and application patterns such as ReAct, with a lab in each of its three modules. We value a clear intermediate scope in 17 hours by Coursera's module cards and a certificate carrying DeepLearning.AI's and AWS's names. What holds the score down is age and depth of practice: Coursera dates its launch to 2023 and states no update since, the labs are guided rather than a project, and agents and current tooling get little time.
3.9 / 5 how well it teaches4.0 / 5 what the certificate is worth
Curriculum currency 3.5 · Completion realism 4.3 · Skill value 3.8 · Employer recognition 4.0 · Cost & value 3.0 · Salary impact 3.4 — the score is the average of these six, each out of five.
Scored with AI assistance against our published rubric; the editor is responsible for the rubric and for every published score.
Check price & enrol on Coursera →
This review covers what each of the course's three modules teaches, what it costs and asks of you, where it has aged, how it compares with the generative AI courses we review, and who should take something else. Everything in it comes from the course's Coursera page as it stood on 8 October 2026.
What is Generative AI with Large Language Models?
It is a single course on Coursera, offered jointly by DeepLearning.AI and Amazon Web Services and taught by four instructors — Chris Fregly, Antje Barth, Shelbee Eigenbrode and Mike Chambers — whom the page describes as expert AWS AI practitioners. Its subject is the life cycle of a generative AI project built on a large language model: choosing and pre-training a model, adapting it by fine-tuning, aligning it with human feedback, and deploying it inside an application.
Coursera lists it as Intermediate, and the description is direct about who it is for: “you should have some experience coding in Python to get the most out of it”, and you should know the basics of machine learning — supervised and unsupervised learning, loss functions, and splitting data into training, validation and test sets. Its FAQ tells anyone without programming experience to start with the Machine Learning Specialization instead.
It is also widely taken: the page showed 449,128 people enrolled when read on 8 October 2026, with an average learner rating of 4.8 from 3,651 reviews. Coursera's page data gives its launch date as 28 June 2023, and the page shows no “Recently updated” line and states no later update date.
What the three modules teach
The course runs as three modules, headed Week 1 to Week 3, each with a quiz and a hands-on lab:
- Week 1: Generative AI use cases, project lifecycle, and model pre-training. The transformer architecture, prompting and generative configuration, the generative AI project life cycle, how models are pre-trained, the computational cost of training and the scaling laws behind compute-optimal models, and pre-training for a specific domain. The lab uses a model to summarise dialogue.
- Week 2: Fine-tuning and evaluating large language models. Instruction fine-tuning on one task and on many, evaluation metrics and benchmarks, and parameter-efficient fine-tuning: LoRA and soft prompts. The lab fine-tunes a model for dialogue summarisation.
- Week 3: Reinforcement learning and LLM-powered applications. Reinforcement learning from human feedback — the reward model, fine-tuning with proximal policy optimisation, reward hacking and scaling human feedback — then optimising a model for deployment and using it in applications: chain-of-thought prompting, program-aided language models, ReAct and LLM application architectures, closing on responsible AI. The lab fine-tunes FLAN-T5 with reinforcement learning to produce more positive summaries.
That is a great deal of the model side of generative AI in a short course, and it is the part most application courses skip: why fine-tuning works, what LoRA changes, what RLHF is optimising and why a reward can be gamed. Someone who finishes can follow a model card or a fine-tuning discussion at work, which is the “practical intuition” the course's own description promises.
Where it has aged is the application layer. ReAct and program-aided language models are as far as it goes towards agents; there is nothing on the Model Context Protocol or on evaluating agents, and the labs are built on FLAN-T5. The ideas transfer; the tooling has moved on since 2023.
Cost, time and prerequisites
Time. About 17 hours by Coursera's module cards. Coursera's headline gives a pace instead, two weeks at ten hours a week. Three weekly quizzes and three labs make up much of it, and the labs are where the learning sticks.
Cost. It can be bought per course or taken through a Coursera Plus subscription, and Coursera prices both by country: its page shows “Included with Coursera Plus” (checked 8 October 2026). Coursera sells Plus by the month or by the year; which is cheaper depends on how long you will study, and Coursera's pricing page shows both plans for your country. Coursera's financial aid is a discount decided course by course, and Coursera says a decision can take up to 16 days.
Labs. The labs run in an environment the course introduces in a video titled “Introduction to AWS labs”, and a reading in the final week is titled “Reminder about end of access to Lab Notebooks”: plan to finish the labs while you are enrolled, and keep a copy of your own work.
Certificate. A shareable Coursera course certificate. The course is offered by DeepLearning.AI and Amazon Web Services, but the certificate is not an AWS certification, and the page names no AWS exam it prepares you for.
Check the price & enrol on Coursera →How we checked this. We list the cost as Paid: bought per course or through a Coursera Plus subscription, priced by country. Source: Coursera course page coursera.org/learn/generative-ai-with-llms, read server-side 2026-10-08: the button reads "Enroll now" and the page shows "Included with Coursera Plus"; it states no price. Coursera prices by country, and this course's enrol screen has not been read in a browser, so no figure is published. Where we give a figure, the source says when we read it on the provider's page. We re-read prices by hand and publish no figure we cannot source — where a provider prices regionally, we say so rather than quote a number that is wrong for most readers.
Pros and cons
✓ What we liked
- Explains how LLMs are pre-trained, fine-tuned and aligned with RLHF
- Covers LoRA and other parameter-efficient fine-tuning
- A lab in every module, ending in reinforcement-learning fine-tuning
- Short and clearly scoped, at about 17 hours
- DeepLearning.AI and AWS on the certificate
✕ What to keep in mind
- Coursera dates it to 2023 and states no update since
- Little on agents, retrieval or current tooling
- Three guided labs rather than a project of your own
- Needs Python and machine-learning basics first
Not sure this is the right one for you?
Tell the picker about your background and what you want the certificate to do, and it narrows the list to the one or two courses we would start with. It suggests only our affiliate partners’ courses, and says so before it suggests anything.
Try the AI Certification Picker →Who should take it, and who should not
Take it if you write some Python, know the machine-learning basics, and want to understand what is going on inside the models your applications call — why fine-tuning helps, when parameter-efficient methods are enough, and what alignment with human feedback does. It also suits product and engineering leads who need to follow conversations about fine-tuning versus prompting with real understanding.
Skip it if you do not code yet: Generative AI for Everyone covers the concepts without code. Skip it, too, if your goal is to build applications — agents, retrieval, tool use — rather than to understand models; DeepLearning.AI's own Agentic AI course is the more current choice for that.
How it compares with the courses we review
The table below compares five generative AI courses on provider, level, time, coding and our rating. Four are Intermediate and use Python; one is a Beginner course with no coding; they run from about 6 hours to 188.
| Course | Provider | Level | Time | Coding | Our rating | Enrol |
|---|---|---|---|---|---|---|
| Generative AI with Large Language Models | DeepLearning.AI & AWS (Coursera) | Intermediate | ~17 hrs | Python | 3.7 / 5 | Coursera → |
| Agentic AI (DeepLearning.AI) | DeepLearning.AI (Coursera) | Intermediate | ~23 hrs | Python | 4.1 / 5 | Coursera → |
| Developing Large Language Models | DataCamp | Intermediate | ~19 hrs | Python | 3.5 / 5 | DataCamp → |
| IBM Generative AI Engineering | IBM (Coursera) | Intermediate | ~188 hrs | Python | 3.8 / 5 | Coursera → |
| Generative AI for Everyone | DeepLearning.AI (Coursera) | Beginner | ~6 hrs | None | 3.1 / 5 | Coursera → |
Agentic AI. DeepLearning.AI's newer course is the application-side counterpart: reflection, tool use and MCP, evaluation and multi-agent patterns, built in Python, with three graded programming assignments. Take this course to understand the models and that one to build with them; if you can take only one and your job is building, take Agentic AI.
DataCamp's Developing Large Language Models. The closest match in subject: a seven-course track on language models in Python, Llama 3, transformers in PyTorch, PyTorch Lightning, RLHF and LLMOps, done as exercises in the browser on a DataCamp subscription. It goes further into code and works with more recent models; this course spends more of its time explaining the ideas, and carries DeepLearning.AI's and AWS's names.
IBM Generative AI Engineering. A sixteen-course Professional Certificate that runs from beginner Python through transformers, fine-tuning, retrieval and LangChain to a project. It is the path to take if you want to build applications and a better-known credential, and it asks many times the hours.
Generative AI for Everyone. The non-technical DeepLearning.AI course on the same subject. It is short and needs no code, but it is bought on its own: Coursera Plus does not include it.
Is Generative AI with Large Language Models worth it?
For the reader it is built for, yes, with one caveat about age. We rate it 3.7 out of 5. It explains the model side of generative AI — pre-training, fine-tuning, parameter-efficient methods and RLHF — in far fewer hours than the longer programmes we review, with a lab in each module and DeepLearning.AI's and AWS's names on the certificate. What holds the score down is that Coursera dates it to 2023 and shows no update since, so agents, retrieval and today's tooling get little time, and its three labs are guided exercises rather than a project you could show an employer.
Take it as the theory course it is. Pair it with a building course — agents or retrieval — and a project of your own, and it will make the second course easier to understand; take it alone expecting a job-ready skill, and it will disappoint.
Ready to start?
Paid through Coursera rather than through the provider, by subscription or per course. Coursera prices by country: its pricing page shows the Coursera Plus plans and the price for your country.
Frequently asked questions
Is Generative AI with Large Language Models worth it?
Yes, for a developer or data scientist who wants to understand the model side of generative AI and accepts that the course dates from 2023. We rate it 3.7 out of 5.
It explains pre-training and scaling laws, instruction fine-tuning, LoRA and other parameter-efficient methods, and reinforcement learning from human feedback, with a lab in each of its three modules, in about 17 hours by Coursera's module cards. That is a lot of the model life cycle in a short course.
What holds it back is age and depth of practice. Coursera states no update since its 2023 launch, so agents and current tooling get little time, and the labs are guided exercises rather than a project of your own. Take it to understand models, and a building course, such as DeepLearning.AI’s Agentic AI, to build with them.
How long does Generative AI with Large Language Models take?
About 17 hours by Coursera's module cards, across three modules headed Week 1 to Week 3. Coursera’s headline gives a pace instead, two weeks at ten hours a week, and its catalogue data suggests a slower one, three weeks at three to four hours a week.
Each module combines videos, readings, a quiz and a lab, and the labs repay working through them properly rather than just running the cells. Treat the hours as a minimum if the machine-learning background is new to you.
Finish the labs while you are enrolled. A reading in the final week reminds learners that access to the lab notebooks ends, so keep a copy of your own work.
Do I need coding experience for Generative AI with LLMs?
Yes, some. The course page says “you should have some experience coding in Python to get the most out of it”, and asks for the basics of machine learning: supervised and unsupervised learning, loss functions, and splitting data into training, validation and test sets.
Its own FAQ is direct about beginners. Asked whether someone with no programming experience can take it, it answers: “We recommend starting with a beginner course such as Machine Learning Specialization.”
Each lab has a walkthrough video, but you need to read Python comfortably to follow what each step of a fine-tuning run is doing. If you want the concepts without code, Generative AI for Everyone is the non-technical option.
Is Generative AI with Large Language Models included in Coursera Plus?
Yes. Its Coursera page shows “Included with Coursera Plus” (checked 8 October 2026), so a Plus subscription covers it; it can also be bought on its own. Coursera prices both by country.
Coursera sells Plus by the month or by the year. Which works out cheaper depends on how long you will study and what else you will take: a learner taking only this course is weighing one purchase against a subscription, while one planning several courses is weighing the two plans. Coursera’s pricing page shows both for your country.
Not every DeepLearning.AI course is in Plus. Generative AI for Everyone and the Agentic AI course are outside it, so check the “Included with” line on each page before you count on a subscription.
Is the course out of date?
Partly. Coursera’s page data gives a launch date of 28 June 2023, and the page shows no “Recently updated” line and states no later update.
The foundations have aged well. How transformers work, how pre-training scales, why instruction fine-tuning and LoRA work, and what reinforcement learning from human feedback optimises are still the ideas behind current models, and the course explains them clearly.
The application layer has aged more. ReAct and program-aided language models are as far as it goes towards agents, there is nothing on the Model Context Protocol or on evaluating agents, and the labs are built on FLAN-T5. Pair it with a current building course if those are what you need.
Does it prepare you for an AWS certification?
Not directly. The course is offered by DeepLearning.AI and Amazon Web Services and taught by instructors the page describes as expert AWS AI practitioners, but the page names no AWS certification exam, and its certificate is a Coursera course certificate, not an AWS credential.
Its subject, how generative AI models are built, adapted and deployed, is useful background for any AI exam. The page does not present it as exam preparation, though, and it should not be used as a substitute for an exam’s own guide.
If an AWS credential is the goal, our AWS Certified AI Practitioner review covers the foundational exam and how to prepare for it.
Keeping this current. Course formats, prices, and certification exam fees change and vary by region. We review our guides regularly, and we always recommend confirming the specifics on the provider's official page before you enrol.