AI jobs salary data shifts every quarter. Real comp shows up in what firms pay, what rivals pay, and what candidates ask for. The hype hides the truth. This guide cuts through it.

The headlines say AI engineers make $1M. That is true for maybe 100 people at OpenAI and Anthropic. For everyone else, the pay picture is more layered. It is also more interesting.

The 2025 compensation picture

Several pay surveys ran in 2024 and 2025. They reached similar results. Levels.fyi gathers real comp from offer letters and signed deals. It shows median total pay for AI and ML engineers at top tech firms ran from $250K to $450K in 2024. Robert Half's 2024 Salary Guide found similar U.S. ranges.

But the top of the market and the median have split apart. Here is where the money actually sits in 2025.

Tier 1: Foundation model researchers ($500K-$5M+)

These researchers build the models at OpenAI, Anthropic, Google DeepMind, xAI, Meta's AI lab, and a few others. New PhD researchers often get offers above $500K total. Senior researchers pass $1M with ease. The very top lead key capability areas. They have reportedly seen offers from $2M to $10M, per 2024 reports from The Information and The Wall Street Journal.

A realism check. This market is real but tiny. There are only a few hundred such jobs worldwide. A PhD plus published research is all but required.

Tier 2: AI infrastructure engineers ($300K-$700K)

These people build and run the systems that AI models train on. The group includes CUDA programmers, distributed systems engineers, GPU cluster operators, and ML platform engineers. They are less famous than the researchers. Yet they are often paid almost as well, since the supply is even tighter. Levels.fyi data from 2024 puts senior infrastructure engineers at top labs from $400K to $700K.

These roles usually need a strong CS background plus distributed systems work. They are hard to enter from outside. Still, they are easier to reach than research.

Tier 3: Senior ML engineers at well-funded startups ($250K-$500K)

This is the applied side. These engineers take a foundation model and turn it into a product. Firms like Perplexity, Cursor, Glean, Harvey, and many other AI-native startups fight for this talent. Pay is high but more shaky. It leans heavily on equity that may or may not match what the cap table suggests.

Required skills here are strong engineering and ease with fast iteration. You must ship products with AI in the loop. An ML or CS degree is not a must. But you do need a portfolio of shipped AI products.

Tier 4: AI engineering at enterprise companies ($150K-$280K)

Banks, insurers, healthcare systems, and retailers are all building internal AI teams. The pay sits below the top labs. Yet the work is interesting and the comp is real. Robert Half's 2024 Salary Guide reports a median of $180K to $220K for senior ML engineers at U.S. enterprise firms.

This is the most reachable high-paying AI engineering market. You do not need to be in San Francisco. The roles spread across all major U.S. business hubs.

Tier 5: AI-adjacent non-engineering ($80K-$200K)

These roles include AI product managers, AI program managers, AI policy specialists, AI sales engineers, and AI consultants. They tap the AI economy without deep engineering skill. Pay is solid but well below the engineering tracks. Indeed Hiring Lab data from 2024 shows median total pay from $100K to $160K for these U.S. roles.

These roles are some of the fastest-growing by raw hire count. They are where the wider AI economy actually employs people.

Tier 6: Prompt engineers and AI content specialists ($60K-$130K)

These were briefly hyped as the next $300K job. The reality is more modest. Indeed and Levels.fyi data from 2024 show prompt engineer median pay from $80K to $130K. That is solid pay, not the unicorn comp the early hype promised.

Here is why. Prompt engineering as a standalone job is folding back into nearby roles. Content strategists, product managers, and ops staff are picking up prompt skills. So the pure prompt engineer role is shrinking.

Where the money is actually moving

Comp data from 2023 to 2025 shows three clear trends:

What pays well that nobody talks about

Here are three high-pay tracks people rarely discuss:

The honest comp advice

Do you want to max out total comp? Then earn a PhD in ML, publish, and get hired into foundation model research. It is a tiny field with huge pay.

Do you want a strong AI career with realistic odds? Then build strong engineering basics, add an AI focus, and ship products that prove the work. After that, pick your region and company stage based on cash or equity.

Do you not want to be an engineer at all? Then the non-engineering AI roles pay well. Not lottery-ticket well, but well above median professional pay.

In all three cases, the pay follows real skills and shipped work. The 2025 market is honest in a way it sometimes has not been. People who can do the work get paid. People who cannot, do not.

Sources

Related reading: What Is an AI Avatar Digital Twin and How Does It Work? · The Best AI SEO Tools (and Their Limits)