State of Tech Hiring 2026

The State of Tech Hiring 2026: what 42,215 direct-apply jobs actually show

I track tech job postings the boring, reliable way: straight from companies' own Greenhouse, Lever and Ashby boards, collected daily and deduped. No aggregator noise, no recycled listings, no ghost jobs that a staffing firm reposts six times. As of 26 June 2026, that dataset holds 42,215 open engineering and product roles across 5,992 companies. This is the direct-apply slice of the market — the jobs you can actually apply to on the employer's site — and it's the cleanest signal I know of for where hiring is really happening.

Here's the headline: the market is moving, not stalling. Over the last 28 days companies opened 16,274 roles and closed 10,890, for a net gain of +5,384. That churn matters more than the static count. Boards that only grow look suspicious; boards where things open and close are alive. And the single most newsworthy pattern in this snapshot isn't a software trend at all — it's hardware. The companies adding jobs fastest right now build rockets, drones, satellites, EVs and the data centers that train AI. More on that below.

Which tech roles are in demand in 2026

By raw volume, two niches own the market and everything else trails.

Full-stack engineering is the biggest single category at 11,249 open roles — roughly one in four jobs I track. It's the default hire when a company needs someone who can ship across the stack, and the breadth of demand reflects that. If you want the widest set of doors, this is it.

Right behind it, AI/ML sits at 9,139 open roles. Together, full-stack and AI/ML account for more than 48% of the entire dataset. That AI number is the one I'd watch hardest: it's the second-largest niche, it pays at the top end, and it's the most remote-friendly category on the board. AI hiring isn't a side bet anymore — it's structural.

After the top two, the volume steps down in a clean line:

• DevOps & Infra — 5,370 open
• Data — 4,251 open
• Product — 3,782 open
• Design — 2,287 open
• Backend — 2,170 open
• Security — 1,732 open
• Mobile — 1,207 open
• Frontend — 1,028 open

One thing worth saying plainly to job-seekers: the gap between full-stack and frontend is roughly eleven to one. Specializing purely in the browser puts you in the smallest pool I measure. That doesn't make frontend a bad bet — it's a discipline, not a job count — but if you're optimizing for the number of openings you can plausibly target, generalist and infrastructure skills give you far more surface area.

Tech salaries 2026: where the money actually is

Pay first, then the big caveat. Only 14% of postings disclose salary at all. Some of that is regulation forcing transparency, some is companies choosing to compete on openness, and a lot of the market still says nothing. So treat every figure below as the disclosed-pay subset, with the sample size (n) attached so you can judge how much weight it carries. I'd trust the full-stack and AI/ML medians far more than the frontend or mobile ones, simply because they rest on hundreds versus a couple dozen data points.

With that out of the way, here's what the disclosing employers pay, by median band:

• Mobile — $155,000–$219,000 (n=28)
• AI/ML — $152,875–$216,061 (n=274)
• Backend — $152,500–$200,000 (n=50)
• Full-Stack — $150,000–$199,880 (n=362)
• Security — $150,000–$179,000 (n=47)
• Product — $145,350–$180,000 (n=133)
• DevOps & Infra — $130,000–$170,000 (n=191)
• Data — $130,000–$167,663 (n=144)
• Frontend — $120,000–$160,000 (n=17)
• Design — $118,240–$166,830 (n=118)

Mobile tops the table on its upper band, but with only 28 disclosed salaries I'd hold that loosely — a handful of senior iOS or Android roles can pull a small sample upward. The number I'd actually stake a claim on is AI/ML: a $152,875–$216,061 median band built on 274 disclosed salaries. That's the rare combination of high pay and a sample large enough to believe. AI/ML isn't just where the second-most jobs are; among well-measured niches, it pays the most.

At the other end, design and frontend anchor the lower medians. Design's range is wide — its top band ($166,830) overlaps with the middle of the pack — which tells you seniority and specialization swing design pay hard. Data and DevOps cluster in the $130K-floor tier despite being high-volume, technical niches; that's worth knowing if you're weighing a pivot purely on comp.

The broad takeaway: across nearly every well-sampled niche, the disclosed midpoint lands in the $150K–$200K corridor. Backend, full-stack, AI/ML and security all live there. The market is paying senior-engineer money as the going rate, not the exception — at least among the employers willing to print a number.

Remote work in 2026: still real, still a minority

Overall, 27% of tracked roles are remote. That's the number to anchor on whenever someone tells you remote is either dead or universal — it's neither. Roughly a quarter of direct-apply tech jobs are location-flexible, and the rest expect you somewhere specific.

Remote share splits cleanly along the software/hardware line. The most remote-friendly niches are AI/ML and Backend at 32%, followed by Product at 31% and Frontend at 29%. These are the disciplines that travel well over a network connection. The least remote? Mobile at 21% and DevOps & Infra at 22%, with Full-Stack at 24% also below the average — pulled down, I suspect, by how many full-stack roles sit inside hardware and on-site-heavy companies.

If remote is non-negotiable for you, the math is simple: aim at AI/ML, backend or product, and accept that you're fishing in the more flexible third of an already-minority pool.

Who's actually hiring: the hardware surge

This is where the dataset earns its keep. Ranked by net new roles over the trailing 28 days — openings minus closings, so we're measuring real expansion, not backlog — the fastest-growing employers are dominated by companies that build physical things.

Two names tie at the top, both adding +367 net roles:

• SpaceX — 1,540 active roles. One of the largest single hiring boards in the entire dataset, and still climbing fast.
• Anduril Industries — 1,652 active roles, the single biggest active board among the movers. Defense hardware is scaling aggressively.

Behind them, the pattern repeats. Shield AI added +132 net (393 active) and ICEYE added +114 net (115 active) — more defense and satellite work. Rocket Lab (+86, 317 active) and drone-delivery firm Zipline (+92, 253 active) round out the aerospace cluster. On the ground, two EV makers are scaling from near-zero visible boards: Rivian (+127 net, 127 active) and Scout Motors (+116 net, 216 active).

The through-line: rockets, drones, satellites and electric vehicles are the loudest hiring stories in this snapshot. If you've been told tech hiring is purely a software-and-AI affair, the movers list says otherwise. The companies adding headcount fastest mostly make hardware — and they need plenty of software, data and infra people to do it. The remote stats and the mover list line up here: hardware companies hire on-site, which is part of why the overall remote share sits at 27% rather than higher.

The AI infrastructure and model layer

The pure-AI names show up too, and they're substantial. OpenAI added +104 net and runs one of the larger boards on the list at 666 active roles. Anthropic added +84 net (365 active). And the standout infrastructure story is Nscale, which added +202 net against 208 active — meaning almost its entire visible board turned over or appeared inside the last month. That's the profile of a company spinning up fast. The compute layer underneath AI is hiring as hard as the model labs themselves.

Fintech and security

Outside hardware and AI, two fintech and security names jump out. Checkout.com added +164 net against 178 active — again, near-total board freshness, the signature of a rapid scale-up. Payments infrastructure firm Plaid added +97 net (99 active), and consumer-security company Gen added +98 net (99 active). Semiconductor-equipment maker ASM rounds out the list at +84 net (280 active), another hardware-adjacent name reinforcing the theme.

What I find most useful about this list is the mix of board sizes. SpaceX and Anduril are adding hundreds of roles on top of already-huge boards — sustained, large-scale hiring. Companies like Rivian, Plaid, Gen and ICEYE show net additions almost equal to their entire active count, which means they've essentially appeared or re-accelerated in the last 28 days. Both are signals worth chasing, but they tell different stories: one is a machine that keeps running, the other is a door that just opened.

What this means if you're job-hunting right now

Let me turn the numbers into a plan.

Play the volume. Full-stack and AI/ML are 48% of everything I track. If you can credibly position into either, you're aiming at the half of the market with the most doors. Frontend and mobile specialists face the thinnest pools — pair the specialty with broader skills to widen your reach.

Follow the pay where the sample is solid. AI/ML offers the strongest combination of high disclosed pay and a large enough sample to trust ($152,875–$216,061, n=274). Backend and full-stack sit right behind in the $150K–$200K corridor. Treat the eye-catching mobile top band cautiously until more employers disclose.

Decide early on remote. A quarter of the market is location-flexible, concentrated in AI/ML, backend and product. Hardware companies — the ones hiring fastest — overwhelmingly want you on-site. You can chase the surge or chase remote, but the data says you'll struggle to do both at once.

Go where the net-new roles are. Don't just scroll big boards; scroll growing ones. SpaceX and Anduril are hiring at scale and steadily. Nscale, Checkout.com, Plaid, Rivian, Gen and ICEYE are scaling so fast their boards are almost entirely fresh — early-mover timing matters there. You can watch this shift in real time on the movers page.

How to read these numbers (and what they don't say)

A few honest limits, because accuracy is the whole point of this report. This is the direct-apply slice — jobs posted on companies' own Greenhouse, Lever and Ashby boards. It is not every tech job in existence; it's the set I can verify and dedupe at the source, which I'd argue is more trustworthy than a bigger, dirtier count. The 42,215 figure is the entire market I track, not an extrapolation.

Salary medians cover only the 14% of postings that disclose pay, and I've shown the sample size for every niche so you can weight them yourself. The net-new figures are a 28-day window, so a single company opening or closing a w