Highest Paying AI Jobs: Real Salaries by Role and Employer

Last Updated: October 4, 2026

The highest paying AI jobs are research scientist and senior engineer roles at frontier labs, where self-reported total compensation runs from about $563,000 to $1.25 million, largely in private-company equity. Filed base pay at OpenAI and Anthropic sits near $300,000, and the US median for computer research scientists is $140,300.

Search for the highest paying AI jobs and you get two different answers. One set of pages lists $150,000 to $350,000 for roles like machine learning engineer, often with no source. Another set of headlines describes researchers offered $100 million.

Neither is useful on its own. Most AI jobs salary data comes from three places (government filings, self-reported total compensation, and press reports), and the figures can differ by orders of magnitude for the same company.

This guide sorts the data by role and by where each figure comes from, so you can see what an AI job pays, which part is cash, and which part is paper. All figures are US data as of October 2026 unless noted, and the type of source is stated each time.

💡 Not sure which AI tool is actually right for your business?

Get the free guide, Which AI Tool Should You Actually Use? — a straightforward breakdown of the leading AI tools to help you pick the right one for your needs.

Subscribe to AI Business Weekly for the guide, plus daily coverage of AI trends, acquisitions, and product launches.

Get the Free Guide →

The highest paying AI jobs are research scientist and senior engineer roles at frontier labs such as OpenAI and Anthropic. Next come senior machine learning engineers at big tech companies and AI product managers. Typical total compensation runs from about $305,000 for product managers to $1.25 million for lab research scientists.

The table ranks the main roles by typical pay and states what each number measures, because the pay types are not interchangeable.

The top rows include stock, while the two government rows are median wages, so the table overstates the cash gap between them.

AI product managers sit in the middle. IdeaPlan puts their median total compensation at $305,000, about 28% above a standard senior product manager, using Levels.fyi, Glassdoor, and PayScale data. That premium suggests the market pays extra for AI experience even outside engineering.

AI salary numbers disagree because they measure different things. Government filings show base pay, crowd-sourced sites show self-reported total compensation including equity, and news reports describe individual offers. The same company can appear as $310,000, $870,000, or $100 million depending on the source.

Three kinds of numbers appear in nearly every AI salary discussion:

Think of the three as a ladder of reliability. Filed base pay is the floor you can check, self-reported total compensation is a plausible middle that moves with stock valuations, and reported packages are outliers. Blending them without saying so is how a single role ends up with a $150,000 floor and a seven-figure ceiling.

OpenAI and Anthropic technical staff earn a median base salary of about $300,000 to $315,000 in federal filings. Self-reported total compensation, which includes equity, runs from roughly $563,000 for senior Anthropic engineers to about $1.25 million for OpenAI research scientists and Anthropic staff engineers, as of mid-2026.

Base pay below comes from H-1B filings, and total compensation comes from self-reported Levels.fyi submissions, as compiled in 2026 by JobsByCulture and HeroHunt.

Most of a senior package at these labs is stock. HeroHunt estimates about 70% of senior packages at OpenAI and Anthropic is illiquid equity rather than cash, so the headline figure depends on what that stock is worth when, and if, it can be sold. Liquidity events like the ones in our guides to the OpenAI IPO and the Anthropic IPO are what turn paper into cash.

An AI software engineer salary at Anthropic starts near $367,000 at entry level in self-reported data, according to HeroHunt, which is well above the enterprise bands in the next section.

For background on the two employers, see our explainers on what OpenAI is and what Anthropic is.

The national median total compensation for AI and machine learning engineers is about $156,000, per Levels.fyi data compiled by HeroHunt in August 2026. Pay rises from $110,000 to $160,000 at entry level to $600,000 or more at principal level, and employer type moves pay as much as seniority does.

The AI engineer salary bands below come from HeroHunt, which cites Levels.fyi data as of August 2026. The same bands describe the machine learning engineer salary picture, and equity makes up a growing share at each step.

The entry level AI engineer salary band is $110,000 to $160,000 in total compensation, about 10% to 20% of it equity.

At senior level, employer type splits the market. Enterprise employers show a median near $245,000, big tech companies like Google and Meta show $393,000 to $502,000, and frontier labs show $600,000 to $1.15 million or more.

That means a senior engineer at a frontier lab reports roughly 2.4 to 4.7 times the pay of a senior engineer at a typical enterprise. Choosing the employer is a bigger lever than choosing the job title.

The mainstream baseline is lower. The Bureau of Labor Statistics lists data scientists at a $120,230 median wage as of May 2025, with employment projected to grow 35% from 2025 to 2035, adding about 95,400 positions. Engineering pay also depends on how AI changes the work itself, which we cover in our analysis of whether AI will replace programmers.

Pay for AI researchers spans a wide range. The US median for computer and information research scientists is $140,300 as of May 2025, while research scientists at Google DeepMind and Meta report roughly $286,000 to $581,000 in total compensation, and OpenAI research scientists report a median near $1.25 million.

The Bureau of Labor Statistics projects 22% growth for computer and information research scientists from 2025 to 2035, and says the typical entry requirement is at least a master's degree in computer science or a related field.

Industry figures sit well above that median. JobsByCulture lists Google DeepMind research scientists at about $286,000 (L4) and $481,000 (L5), and Meta research scientists at about $305,000 (IC4) rising to $581,000 (IC6). Filed base pay is far more modest: HeroHunt's review of Meta's H-1B filings shows research scientist base salaries of $164,000 to $302,000, with a median near $182,000.

The title varies (AI researcher, AI scientist, research scientist), but pay depends mainly on the employer. The gap in AI researcher salary figures reflects two labor markets.

One covers research roles across government, universities, and companies. The other is a few well-funded labs competing for people who can build frontier models.

They are reported, not confirmed. Wired reported that Meta offered top AI researchers up to $300 million over four years, and OpenAI's Sam Altman cited $100 million signing bonuses. Meta's CTO called that characterization dishonest, and the figures describe a handful of named individuals rather than a market rate.

Axios summarized Wired's July 2025 reporting that offers could reach $300 million over four years, with more than $100 million in the first year. Altman told OpenAI staff Meta had made "giant offers to a lot of people on our team," including $100 million signing bonuses.

Meta's CTO Andrew Bosworth pushed back at an all-hands meeting. "Sam is just being dishonest here," he said, per Entrepreneur. He added: "Look, you guys, the market's hot. It's not that hot."

HeroHunt compiles other reported figures, including up to $250 million for researcher Matt Deitke and more than $200 million for Ruoming Pang, and notes these are multi-year stock packages rather than cash bonuses.

For job seekers, these numbers are a signal of how hard a few labs compete for a few people, not a salary benchmark. Ordinary senior roles at Meta look different: HeroHunt lists E5 engineers near $424,000 and E6 near $687,000, with E6 machine learning engineers near $763,000. Our explainer on Meta AI covers what the company is building.

AI skills carry a wage premium, but the size depends on the study. PwC's 2026 barometer found 62% across 27 countries and territories, up from 57% a year earlier. Lightcast's analysis of 1.3 billion postings found 28%, nearly $18,000 a year, in 2025.

PwC's 2026 Global AI Jobs Barometer analyzed more than one billion job ads. It found the premium ranges from 16% in government to 118% in consumer markets, and that AI-specific roles grew 69%, about eight times faster than the 9% growth of the overall jobs market.

Lightcast compared postings that list AI skills with those that do not. Its release does not specify geography.

Both studies measure advertised pay, not what a worker earns after learning a skill. Use them to see where demand pays, not to forecast a personal raise. Our AI job market statistics page tracks the hiring data behind these trends.

Level and employer type decide pay more than job title. Research roles typically require at least a master's degree, per the Bureau of Labor Statistics, while engineering pay rises with seniority and with moving from enterprise employers to big tech or frontier labs. Government projections show strong growth: 22% for research scientists and 35% for data scientists through 2035.

The pay data above points to three levers. Seniority matters, since principal-level engineers report $600,000 or more against $110,000 to $160,000 at entry level. Employer type matters, since frontier labs and big tech sit well above enterprise at the same level.

Role matters too, since AI product managers report a 28% premium over standard senior product managers.

Entry-level hiring is a separate challenge from pay, and our guide to AI and entry-level jobs for college graduates covers it. The salary figures in this guide mostly describe experienced hires.

Research scientists and senior engineers at frontier labs earn the most. Self-reported Levels.fyi data compiled by HeroHunt puts OpenAI research scientists near $1.25 million in median total compensation, though much of that is private-company equity, while filed base pay at the same labs is closer to $300,000.

Pay at the very top is concentrated in a small group of employers, which our list of the biggest AI companies shows.

OpenAI technical staff show a median base salary of about $310,000 in Department of Labor H-1B filings. Self-reported total compensation including equity runs from roughly $635,000 to $653,000 at mid level and $843,000 to $1.16 million at senior level, depending on the snapshot.

An OpenAI salary is mostly stock at senior levels, so the cash portion is far smaller than the total. See our explainer on what OpenAI is for the company background.

The national median total compensation for AI and machine learning engineers is about $156,000, according to Levels.fyi data compiled by HeroHunt in August 2026. Bands rise from $110,000 to $160,000 at entry level to $600,000 or more at principal level.

Employer type matters as much as seniority, with senior frontier-lab engineers reporting $600,000 to $1.15 million or more.

The US Bureau of Labor Statistics puts the median for computer and information research scientists at $140,300 as of May 2025. Research scientists at Google DeepMind and Meta report about $286,000 to $581,000 in total compensation, and OpenAI research scientists report a median near $1.25 million.

An AI research scientist salary depends heavily on whether the role is at a frontier lab or in a government, university, or corporate research setting.

Yes, but the estimates vary widely by study. PwC's 2026 Global AI Jobs Barometer found a 62% wage premium across 27 countries and territories, while Lightcast's 2025 analysis of 1.3 billion postings found 28%. Both compare advertised pay, not individual raises.

For the hiring data behind these figures, see our AI job market statistics.

For research roles, usually yes. The Bureau of Labor Statistics says computer and information research scientists typically need at least a master's degree in computer science or a related field. Engineering and product roles are less uniform, and this guide's pay data is organized by role and level rather than degree.

Research-track roles are the most degree-dependent. For the entry-level side, see our guide to AI and entry-level jobs for college graduates.

It is a press report, not a confirmed salary. Wired reported Meta offers of up to $300 million over four years, as summarized by Axios, and OpenAI's Sam Altman cited $100 million signing bonuses. Meta's CTO Andrew Bosworth called that framing dishonest, saying "the market's hot. It's not that hot."

These figures describe a handful of individuals and are mostly multi-year stock packages. Our explainer on Meta AI covers the company behind the offers.

The highest paying AI jobs are research scientist and senior engineer roles at frontier labs, where self-reported total compensation runs from about $563,000 to $1.25 million, much of it private-company equity. Filed base pay at OpenAI and Anthropic is about $300,000, senior AI engineers at big tech earn $393,000 to $502,000, AI product managers earn about $305,000, and the US median for computer research scientists is $140,300. Reported nine-figure offers at Meta describe a handful of individuals, not the market. All figures are US data as of October 2026 and should be rechecked before a salary negotiation.

By Sameer Khan

AI-assisted. Researched, reviewed, and edited by Sameer Khan before publication.