AI certifications from Google, Microsoft, AWS, NVIDIA and more: which are free?
By Xplaind editors ·

Every big technology company now offers AI training, and most of it is free. The proctored exams that lead to a formal certification are not. This guide separates the two, so you can spend money only where it helps.
All fees below were checked on the official exam pages on 25 October 2026. Prices change and taxes are extra, so confirm on the exam page before you book. Our certifications page keeps the same list up to date with links to every exam and its free prep.
Paid exams worth knowing
| Certification | Provider | For | Exam fee (checked 25 Oct 2026) |
|---|---|---|---|
| AWS Certified AI Practitioner (AIF-C01) | Amazon Web Services | Anyone working with AI on AWS, technical or not | US$100 |
| Azure AI Fundamentals (AI-901) | Microsoft | Early-career AI developers on Azure | Set by country |
| Azure AI Apps and Agents Developer Associate (AI-103) | Microsoft | Developers building AI apps and agents on Azure | Set by country |
| Generative AI Leader | Google Cloud | Managers and non-engineers leading AI work | US$99 |
| Professional Machine Learning Engineer | Google Cloud | Experienced ML engineers | US$200 |
| NVIDIA-Certified Associate: Generative AI LLMs (NCA-GENL) | NVIDIA | Developers new to generative AI | US$125 |
| Generative AI Engineer Associate | Databricks | Engineers building RAG and agents on Databricks | US$200 |
Two changes in 2026 are worth knowing. Microsoft retired its long-running AI-900 exam; the Azure AI Fundamentals certification now uses AI-901. And the developer exam AI-102 was replaced by AI-103, which is built around generative AI and agentic solutions in Microsoft Foundry. Older blog posts and videos still describe the retired exams.
Which one should you take?
- You are not an engineer (sales, operations, product, management): AWS AI Practitioner or Google Cloud Generative AI Leader. Both test understanding, not coding.
- You are starting a technical AI career: Azure AI Fundamentals if your target employers use Microsoft, otherwise AWS AI Practitioner. Then build projects.
- You build AI applications already: the cloud developer exams (Microsoft AI-103, Databricks Generative AI Engineer) or NVIDIA's NCA-GENL, which tests the fundamentals behind language models rather than one vendor's tools.
- You are an experienced ML engineer: Google Cloud Professional Machine Learning Engineer.
In India, many IT services companies and cloud partners value these exams because they count towards partner status. Product companies and startups care far more about what you have built.
Free courses that end in a certificate or badge
These cost nothing. They show you finished a course, which is useful for structure and for your LinkedIn profile, but they are not proctored exams.
- OpenAI Academy. Free tracks for work, building, leading AI adoption and teaching, each with a course badge when you pass the assessment. Start with Apply AI at Work.
- Anthropic Academy. Free courses with certificates of completion, from Claude 101 for everyday use to Building with the Claude API for developers, plus courses on agent skills and the Model Context Protocol.
- Google's video lessons. The Introduction to Generative AI learning path is a short, beginner-friendly set of video lessons. Google gives skill badges for some paths, but they are not certifications.
- Databricks Academy. Generative AI Fundamentals is a short course with a badge; Databricks has made its self-paced training free.
- DeepLearning.AI. AI For Everyone and AI Prompting for Everyone have free video lessons; the graded assignments and certificate need a paid plan.
Microsoft's own Generative AI for Beginners and AI Agents for Beginners courses are free and excellent, but they are open courses on GitHub without a certificate.
How to prepare for an exam without paying for a course
- Download the official exam guide or study guide and list every topic it covers.
- Take the provider's free learning path. Every exam above has one; our certifications page links them.
- Fill the gaps with free courses. Our course directory lets you filter by topic and by "certificate or badge".
- Do hands-on labs on a free tier, and set a spending alert first.
- Take the official practice assessment where one is free, such as Microsoft's practice assessments.
The honest summary
If you can afford only one thing, spend your time on a project rather than your money on an exam. If a job you want asks for a certification, pick the one for that employer's cloud, prepare with the free path, and book the exam when you pass practice tests comfortably.
Questions people ask
- Are there free AI certifications?
- Free certificates of completion and badges, yes: OpenAI Academy, Anthropic Academy and Databricks Academy offer free courses that end in a certificate or badge, and Google's video lessons are free to watch. Proctored certification exams from AWS, Microsoft, Google Cloud, NVIDIA and Databricks cost money, but all of them can be prepared for with free courses.
- Which AI certification is best for a beginner in India?
- For non-engineers, AWS Certified AI Practitioner or Google Cloud Generative AI Leader. For people starting a technical AI career on Azure, Microsoft's Azure AI Fundamentals (exam AI-901). Choose the cloud your target employers use.
- Do AI certifications help you get a job?
- They help when a job description asks for one, especially at IT services companies and cloud partners, and they give your learning structure. They do not replace projects: interviewers will ask what you built.