The State of Remote Engineering Work in 2026

The argument over remote work in software engineering is no longer theoretical. Two years after the initial wave of return-to-office mandates, the dust has settled enough to see what the market actually looks like. We analyzed 77,480 active software engineering job listings across 214 countries and 8,230 companies to find out where the industry has landed -- and the picture is more fractured than either side of the debate tends to admit.

The headline numbers tell a story of a market that has found an uneasy equilibrium rather than a decisive winner. Onsite work has reclaimed the plurality, but it has not reclaimed dominance. Remote work has contracted from its pandemic peak, but it has not collapsed. And hybrid -- that awkward middle child of workplace policy -- has quietly grown into the fastest-expanding category, driven by companies that want employees nearby but cannot quite justify requiring them at a desk five days a week.

What makes these numbers worth examining closely is not the top-line split. It is the variation underneath it. The question is no longer "is remote work dead?" but rather: who gets to work remotely, in which industries, at which levels, and at what price? The answers reveal a job market that is stratifying along new lines -- where your seniority, your specialization, and the industry you work in determine your workplace flexibility as much as any company-wide policy does.

The Three-Way Split

Of the 77,480 software engineering positions in our index, 33,796 require full-time onsite presence. That is 44% of the market -- a substantial share, and one that has grown from an estimated 30% in early 2024. The return-to-office movement has real teeth: nearly half of all engineering jobs now expect you in a specific physical location every working day.

But those numbers obscure important nuance. The remaining 2% of listings -- roughly 1,954 jobs -- either do not specify a remote policy or use ambiguous language ("flexible", "sometimes remote") that does not map cleanly to any of the three categories. We exclude these from the percentage calculations that follow, but their existence is a reminder that workplace policy remains a messy, inconsistently communicated thing.

The more important takeaway is that remote work, while smaller than it was at its peak, has stabilized. The 29% figure has held remarkably steady over the past two quarters. The companies that went remote have largely stayed remote. The return-to-office mandates that were going to happen have mostly already happened. What we are looking at now is less a trend in motion and more a new equilibrium -- one where roughly three in ten engineering jobs do not care where you live.

Which Industries Lean Remote?

The aggregate split tells you what the average software engineering job looks like. But nobody works at the average company. The industry you target changes your odds dramatically.

Developer tools companies are the most remote-friendly segment in our data, with approximately 45% of their listings offering fully remote work. This is not surprising. DevTools companies sell to distributed engineering teams and have strong cultural incentives to practice what they preach. Their products are built for asynchronous collaboration, and their workforces tend to reflect that philosophy.

Remote share by industry

At the other end of the spectrum, defense sits at roughly 8% remote. Security clearance requirements, classified networks, and physical facility constraints make remote work structurally difficult in that sector, regardless of management philosophy. Healthcare and insurance, both undergoing massive legacy system modernization, also skew heavily onsite -- likely because the complexity of integrating with on-premise hospital systems and insurance mainframes requires physical proximity to infrastructure that cannot easily be replicated in a home office.

The middle of the distribution is where the most interesting story lives. Fintech -- the largest hiring sector in our index at 9,905 listings -- sits almost exactly at the market average of 29% remote. This reflects the tension within financial technology between the move-fast culture of startups (which tend to be remote-friendly) and the regulatory-compliance culture inherited from banking (which tends to require butts in seats, especially for roles touching sensitive financial data).

AI and machine learning companies clock in at approximately 40% remote, the second-highest of any industry. The talent market for ML engineers is so competitive that companies in this space have less negotiating power over location requirements. When there are five recruiters in your inbox for every ML engineer, insisting on a San Francisco office is a luxury many cannot afford.

Seniority and Remote Access

Perhaps the most striking pattern in the data is the correlation between seniority and remote availability. Put simply: the more experienced you are, the more likely you are to find remote work. This is not a gentle slope. It is a staircase.

Remote availability by seniority level

At the junior level, only 22% of listings are remote. At staff and principal level, that figure climbs to 38% -- nearly double. The reasons are both economic and practical. Senior engineers have more leverage in the job market. They have established track records that reduce the perceived risk of unsupervised remote work. And the work they do -- system design, architecture, code review, mentorship -- often lends itself to asynchronous communication more naturally than the task-driven, closely supervised work that many junior roles entail.

There is also a selection effect at play. Companies that offer remote staff-level positions tend to be the ones that have built the organizational infrastructure to support distributed teams: strong documentation practices, async-first communication norms, clear goal-setting frameworks. These are also, not coincidentally, the companies that tend to be most attractive to experienced engineers. The result is a virtuous cycle where the best remote companies attract the best senior talent, which in turn makes remote work at that level function better, which makes more companies willing to try it.

For junior engineers, the implication is sobering. The remote-first world that many entered the industry expecting has narrowed considerably for their cohort. This does not mean junior remote jobs do not exist -- our index contains hundreds of them -- but competition for those positions is substantially fiercer than for their onsite equivalents.

The Remote Salary Premium

One of the most persistent questions in the remote work debate is whether distributed engineers earn more or less than their onsite peers. The data provides a clear answer, at least at the senior level: remote pays more.

Senior remote software engineers in our dataset earn an average of $191,000, compared to $180,000 for their onsite counterparts -- a 6% premium, or roughly $11,000 per year. This is not a small number. Over a five-year career stretch, it compounds to more than $55,000 in additional gross earnings, and that is before accounting for the commuting costs, wardrobe expenses, and lunch spending that onsite work typically entails.

The premium exists for several reasons. First, remote roles are disproportionately posted by well-funded technology companies -- the DevTools, AI/ML, and SaaS firms that also happen to pay the highest salaries. Second, companies offering remote positions are competing in a national (or global) talent pool, which means they have to price their roles against the highest-cost-of-living markets even if the engineer lives somewhere cheaper. A remote role listed by a company headquartered in Boise is still competing with remote roles from companies in San Francisco, and the salary has to reflect that.

Third, and perhaps most importantly, remote senior roles tend to require a higher degree of autonomy and self-direction than equivalent onsite roles. The premium is partly compensation for a skill that is difficult to screen for and genuinely scarce: the ability to deliver consistently without the scaffolding of an office environment. Not every senior engineer can do this effectively, and the ones who can are paid accordingly.

It is worth noting that this premium is specific to the senior-and-above level. At the junior and mid-level, the salary difference between remote and onsite is negligible and inconsistent across industries. The market has not yet decided that junior remote work is valuable enough to pay extra for. Whether that changes as a generation of engineers raised on remote collaboration enters mid-career remains to be seen.

Geographic Patterns

Fully remote jobs are, by definition, location-independent. But that does not mean they are evenly distributed. When we look at where remote jobs are posted (based on the company's stated headquarters or primary hiring location), clear geographic clusters emerge.

Top cities for remote engineering roles

San Francisco leads with 3,200 remote engineering listings, followed by New York at 2,100 and Austin at 900. This might seem paradoxical -- why would the most expensive city in the country be the hub for jobs that do not require you to live there? The answer is that "remote" in job listings often means "remote, but headquartered here." The company culture, the time zone overlap expectations, and sometimes even the salary band are still anchored to the headquarters location.

Austin's emergence as the third-largest remote engineering hub reflects a broader trend. The city has attracted a wave of technology companies over the past five years -- some relocating from the Bay Area, others expanding their footprints -- and many of these companies brought remote-friendly cultures with them. Austin-headquartered remote roles offer an interesting middle ground: competitive salaries benchmarked to a high-cost-of-living city, but often without the San Francisco premium.

Outside the US, London stands out with 580 remote listings, but the European remote market is structurally different. European remote roles are more frequently constrained to specific time zones or EU member states, reflecting the complexity of cross-border employment law and tax obligations. A "remote" listing from a London company is far more likely to mean "remote within the UK and Ireland" than truly global remote.

Remote Work by Technology

The technology you specialize in also affects your remote prospects. Cloud infrastructure and platform engineering roles -- the kind that involve Kubernetes, Terraform, and AWS -- are the most remote-friendly technical specializations. This makes intuitive sense: if your job is building and maintaining systems that exist entirely in the cloud, the argument for physical co-location is thin.

On the other end, mobile development (iOS, Android) and embedded systems roles are the least likely to be remote. Embedded work often requires access to physical hardware, test rigs, and lab equipment. Mobile development, while not inherently location-dependent, tends to be done at larger consumer-facing companies that are more likely to have returned to onsite work.

Full-stack and backend web development roles fall near the market average. Frontend-specific roles are slightly more remote-friendly than backend, likely because frontend work tends to be more modular and less dependent on access to production infrastructure. But the differences within web development are small compared to the gulf between cloud/infra and embedded.

The ATS Landscape

One subtle but revealing signal comes from the applicant tracking systems companies use. Our index pulls from 21 ATS platforms, and the distribution is not even. Greenhouse leads with 19,581 listings, followed closely by SmartRecruiters at 19,412. Ashby, a newer entrant popular with startups, contributes 10,139 listings, while Lever accounts for 9,737.

Why does this matter for remote work? Because the choice of ATS correlates with company stage, culture, and hiring philosophy. Ashby-using companies, for example, have a meaningfully higher remote rate than the market average -- consistent with their user base of venture-backed startups that often began as remote-first organizations. Workday-using companies, which tend to be larger enterprises, skew heavily onsite. The ATS a company uses is not causal, but it is a surprisingly useful proxy for how that company thinks about work.

What Comes Next

Predicting the future of remote work in software engineering is a good way to be wrong. But the data suggests several trends worth watching.

Hybrid will keep growing, mostly at the expense of onsite. Companies that mandated five-day RTO and lost talent are quietly walking it back to three or four days. The ones that held firm are finding it harder to fill roles. Hybrid is the compromise that lets companies claim they have an office culture without requiring it to be full-time, and that compromise is attractive to a lot of organizations that are tired of fighting about it.

Remote will stratify further by seniority. The gap between junior and senior remote availability is already large and likely to widen. As companies develop better remote management practices, those practices will be reserved for the experienced engineers who need the least management. Junior engineers will continue to face pressure to be onsite, particularly at companies that view in-office time as a training and mentorship mechanism.

The geographic anchoring of remote roles will loosen slowly. Today, most remote jobs are still listed with a headquarters city and often carry implicit time zone expectations. As distributed team tooling matures and companies gain more experience managing across time zones, expect to see more truly location-agnostic listings -- but the change will be gradual, not sudden.

Compensation models will diverge. Some companies will continue to pay location-adjusted salaries for remote workers. Others will adopt single-band compensation where everyone at a given level earns the same regardless of where they live. Both approaches have trade-offs, and neither has clearly won. The 6% remote premium we observe today is an average that masks significant variation, and that variation will likely increase before it narrows.

The most important thing the data reveals is that remote work in software engineering is not a temporary experiment or a fad that is ending. It is a permanent structural feature of the market that varies dramatically by industry, seniority, technology, and geography. The question for individual engineers is not whether remote work exists, but whether it exists for the specific kind of work they do, at the level they are at, in the industry they want to be in. For some, the answer is overwhelmingly yes. For others, the path to remote work requires deliberate career positioning. The numbers show both realities with equal clarity.

Methodology

Related Research