CertifyDeckAn AI engineer earns $187,700 a year on average in the United States. In Spain it is €36,000 to €60,000 gross, depending on whether you build applications on existing models or do core machine learning; in Mexico the reference is 19,400 pesos a month and in Colombia 6.9 million. Of the five profiles we have analyzed, it is the best paid by a distance.
AI engineer salary by country
The job has two names in postings, AI engineer and machine learning engineer, and in Mexico and Colombia it almost always appears as data scientist.
| Country | Role | Average salary | Usual range |
|---|---|---|---|
| United States | Machine learning engineer | $187,675 a year | $111,186 – $316,783 |
| Spain | Machine learning engineer | €59,710 a year | €34,823 – €102,383 |
| Spain | AI engineer | €36,286 a year | €28,976 – €45,442 |
| Mexico | Data scientist | MXN 19,392 a month | MXN 12,601 – 29,842 |
| Colombia | Data scientist | COP 6,909,154 a month | COP 3,859,617 – 12,368,172 |
Indeed: United States, "Machine learning engineer" (about 5,200 salaries, 28 Sep 2026); Spain, "Machine learning engineer" (26, 28 Sep 2026) and "AI engineer" (5, 27 Sep 2026); Mexico, "Data scientist" (100, 17 Sep 2026); Colombia, "Data scientist" (24, 25 Sep 2026).
What postings pay today
There were 262 artificial intelligence postings live today on Tecnoempleo, the Spanish technology job portal. The ones that state a salary draw the ladder clearly:
For someone coming from software development and starting to build applications with generative AI, our estimate in Spain is €35,000 to €45,000 for the first couple of years.
| Role | Location | Gross annual salary |
|---|---|---|
| Data & AI Engineer | Madrid, hybrid | €30,000 – €42,000 |
| AI Senior Engineer | Seville | €45,000 – €54,000 |
| Senior AI Engineer | Madrid, hybrid | €54,000 – €60,000 |
| AI Product Owner Engineer | Barcelona, hybrid | €60,000 – €72,000 |
Tecnoempleo postings on 2 October 2026. The €35,000–45,000 range is a CertifyDeck estimate based on Indeed's AI engineer average and these postings.
Two different jobs under one name
"AI engineer" covers two profiles. One trains and deploys its own models: it needs statistics, deep Python and experience with data at scale, and it takes the salaries at the top of the table. The other builds products on models that already exist: it connects a language model to company data, builds an agent that carries out tasks, evaluates the answers and watches the cost.
The second is the one you reach soonest from software development. It does not require a PhD, but it does require knowing retrieval, evaluation, content safety and running things in production.
Where AI-103 fits
AI-103 (Azure AI Apps and Agents Developer Associate) is designed for that second profile. The largest block of the exam is implementing generative AI and agentic solutions (30–35%), followed by planning and managing the solution on Azure (25–30%); vision, text analysis and information extraction share the rest (10–15% each).
It does not turn anyone into a machine learning engineer or promise a salary. It shows you can take an AI application to production on Azure, which is what the entry-level postings in the table ask for.
If you are thinking about taking AI-103
First check how much you already know. On CertifyDeck you can start free, with no card, with AI-103 questions and an explanation on each. Go to the AI-103 practice questions.
If you are not sure it is worth it, we go through it in is AI-103 worth it?, and the study plan is in how to pass AI-103.
Frequently asked questions
How much does an AI engineer make in the US?
$187,675 a year on average for a machine learning engineer, ranging from $111,186 to $316,783, according to Indeed (about 5,200 salaries, September 2026).
How much does an AI engineer make in Spain?
Between €36,000 and €60,000 gross a year on average depending on the profile: €36,286 for AI engineer and €59,710 for machine learning engineer, according to Indeed. Senior postings in October 2026 reach €54,000–60,000.
Do you need a master's degree to work in AI?
Training your own models usually requires a strong background in statistics and machine learning. Building applications and agents on existing models is reachable from software development, and it is the profile AI-103 covers.
Where do the figures come from?
From Indeed's salary pages for each country (salaries published in job postings), the US Bureau of Labor Statistics Occupational Outlook Handbook and Tecnoempleo postings with a salary range, checked on 2 October 2026. In Mexico and Colombia the reference role is data scientist. They are gross amounts, excluding bonuses and benefits.