Meta’s hiring push in 2025 offered a revealing look at how fiercely the technology industry is competing for artificial intelligence talent. Federal work visa filings show the company was willing to offer exceptionally high base salaries across engineering, product, research, and leadership roles, with the richest packages concentrated in machine learning, infrastructure, and advanced AI research.
The figures are striking not only because of their size, but because they capture a broader shift in how major tech firms now value technical expertise. In a market defined by generative AI, cloud scale computing, and model development, companies are no longer simply hiring software workers. They are bidding aggressively for specialists who can help define the next wave of products and infrastructure. Meta’s compensation data makes that competition unusually visible.
The salary disclosures also show the widening gap between standard technology hiring and the elite end of the AI labor market. While many employees still fall into the upper six figure range typical of large U.S. technology firms, a smaller group of highly specialized workers is being paid at levels that reflect not just skill, but scarcity and strategic importance.
Meta paid heavily for technical and AI roles
According to an analysis of more than 5,800 federal H-1B and other work visa applications filed by Meta in 2025, most of the company’s new hires earned between $150,000 and $250,000 in base salary. But the upper end of the pay scale was far more dramatic. One software engineer was offered a base salary of $450,000, a research engineer reached $400,000, and a product manager received $348,101 before stock awards, bonuses, and other benefits were added.
Those numbers show how much Meta was prepared to spend to secure in demand technical talent. About half of all the jobs filled through the H-1B program during the period were in software engineering, confirming that engineering remains the core of the company’s hiring strategy. But the data also makes clear that not all engineers are valued equally. Machine learning, AI systems, and senior infrastructure roles sit at the top of the compensation ladder.
The base salaries alone are substantial, but they still understate the real economics of high end tech recruiting. These disclosures exclude stock grants, signing bonuses, and additional perks, which can dramatically increase total compensation and make the most competitive packages far more expensive than the salary line suggests.
AI specialization commands the strongest premiums
The clearest pattern in the data is the premium attached to artificial intelligence related expertise. AI research scientists were paid base salaries ranging from $163,800 to $328,000. Machine learning engineers earned from $165,000 to $250,602, while software engineers focused on machine learning ranged from $144,096 to $293,118. Other ML oriented roles, including systems engineers and managers, regularly pushed into the high $200,000s.
At the senior end, the numbers became even more revealing. Meta’s vice president of engineering for AI earned a base salary of $650,000, making it one of the clearest examples of how central AI leadership has become to the company’s strategy. The company has reportedly offered total compensation packages above $100 million to top researchers, illustrating that the most prized talent is being pursued with a level of financial intensity more commonly associated with star executives than technical staff.
This is the real message inside the filings: AI is no longer just one area of recruitment among many. It is a priority category where Meta is willing to stretch compensation structures to levels that signal both urgency and long term strategic dependence.
Research, product and data roles also saw strong pay
The hiring spree was not limited to pure engineering positions. Research scientists, UX researchers, data scientists, product designers, and technical program managers were also paid at levels that reflect a broad competition for multidisciplinary talent. Research scientist salaries ranged as high as $302,134, while UX research roles climbed to $292,160 and research science managers exceeded $300,000.
In the data organization, data scientists reached base salaries of nearly $296,000, with data engineering managers and analytics leaders regularly moving above $280,000. Product roles also stood out. Product managers were paid as much as $348,101, product design directors exceeded $340,000, and technical program managers approached $290,000. These figures show that Meta’s hiring push was not only about model builders. It also relied on people who could shape products, manage large systems, and turn research into deployable services.
The result is a compensation picture that reflects how modern AI development actually works. It is not driven by a single role type, but by a network of engineers, researchers, designers, analysts, and managers whose work must fit together across large and complex product environments.
Hiring remained aggressive despite policy and restructuring pressure
Meta’s compensation push came during a year in which the broader hiring environment was becoming more complicated. Business Insider reported that the company’s H-1B visa filings roughly halved in the final quarter of 2025 compared with the same period a year earlier. That decline coincided with changes introduced by the Trump administration in September that made the visa process more expensive and subjected applications to closer scrutiny.
At the same time, Meta was still actively hiring while also cutting staff in parts of the business such as Reality Labs. That contrast shows how selective the company’s labor strategy has become. It is reducing headcount in less central or less efficient units while continuing to spend aggressively on the kinds of workers viewed as essential to the next phase of competition in AI.
By the end of 2025, Meta had 78,865 employees, according to its annual filing. The latest salary disclosures suggest that within that workforce, a growing share of value is being concentrated in highly specialized technical roles. In that sense, the filings are about more than pay. They show how the company is reorganizing itself around artificial intelligence, and how expensive it has become to secure the people expected to build that future.