AI/ML Jobs Hiring Report — July 2026

Key facts

• As of July 2026 there are 9,819 open AI/ML roles across 3,121 companies in the direct-apply dataset.
• Net hiring is positive but thin: +479 net new roles over the last 28 days (3,659 opened against 3,180 closed).
• 569 companies posted their first-ever AI/ML role in the last 28 days — new demand, not just churn from the usual names.
• Median disclosed pay sits at $158,797–$214,781/yr, but only 18% of postings (n=295) name a number.
• Weloglobal added the most, +104 net roles in 28 days — more than double the next company.

The single most striking thing in this month's pull isn't the headline count — it's that one company, Weloglobal, opened 104 net AI/ML roles in 28 days and now accounts for its entire active board. That's more than the next two fastest-hiring companies combined. When a single employer moves that hard in a month where the whole market only netted +479, it distorts the picture, and you should read the aggregate numbers with that in mind.

How many AI/ML jobs are open right now?

There are 9,819 open AI/ML roles across 3,121 companies as of this July 2026 snapshot. That's a live count from companies' own Greenhouse, Lever and Ashby boards, so it's postings people can actually apply to directly — not aggregator noise, reposts, or staffing-agency listings.

Breadth is the story here more than depth. 3,121 companies for 9,819 roles works out to roughly three openings each on average, and the median company is running one or two. This isn't a market where a handful of labs hoard all the hiring; it's spread wide, with most employers dipping in for a role or two at a time. You can browse the full set on the AI/ML jobs page.

Is AI/ML hiring growing or cooling?

It's growing, but barely, and the growth is lumpier than the +479 net figure suggests. Over the last 28 days, 3,659 roles opened and 3,180 closed. That's a lot of movement on both sides — churn is high, which is normal for a category where companies post, fill or pull roles quickly. The net is positive, so on balance more doors are opening than closing.

Strip out Weloglobal's +104 and the rest of the market netted around +375 — still positive, still healthy, but you can see how much a single outlier tilts the read. The number I'd actually anchor on is the 569 companies that posted their first AI/ML role in the last 28 days. New entrants are a better signal of genuine demand expansion than net counts, because they represent employers who weren't in this category at all a month ago and now are. That's the part of the data I'd call encouraging. Track the month-to-month movement on the AI/ML movers page if you want to watch it yourself.

What do AI/ML jobs pay?

The median disclosed band is $158,797–$214,781/yr, but the honest headline is that only 18% of postings disclose pay at all — 295 out of the full set. So treat that band as directional, weighted toward US roles and toward employers in states or cities that legally require salary ranges. The other 82% are a blank.

Within the roles that do post numbers, the three big titles land close together:

• Machine Learning Engineer: $159,300–$230,000 (n=41)
• AI Engineer: $159,300–$214,500 (n=32)
• Research Scientist: $150,000–$233,950 (n=26)

Two things worth pulling out. Research Scientist has the lowest floor but the highest ceiling of the three — that spread reflects how much the title stretches, from someone a year out of a PhD to a staff-level researcher. And ML Engineer and AI Engineer floors are basically identical at $159,300, which tells you the market isn't yet paying a clear premium for one label over the other; the distinction between "AI Engineer" and "ML Engineer" is still mushy in a lot of job descriptions. Sample sizes are small (26 to 41 roles each), so don't over-index on the exact numbers. More detail sits on the AI/ML salaries page.

Which roles and seniority levels dominate?

Engineering, overwhelmingly. Of 9,819 roles, 9,366 are engineering — north of 95%. Data analytics is a distant second at 354, and everything else (operations, design, product, marketing) barely registers in single or double digits. If you're a non-engineer hoping to break into AI/ML through this dataset, the honest read is that the direct-apply market is almost entirely building roles right now. Product and design AI hiring is happening somewhere, but it's not showing up here in volume.

On seniority, the biggest bucket is "unspecified" at 4,818 — nearly half. That's a data limitation as much as a finding: plenty of boards don't tag a level, and I won't pretend to infer one that isn't stated. Of the roles that do specify:

• Senior: 2,800
• Principal: 1,342
• Director: 372
• Junior: 351
• Executive: 105
• VP: 31

Here's the number that should stop you: there are 1,342 principal-level roles and only 351 junior ones. Principal openings outnumber junior nearly four to one. That's a market that wants people who've already done the work — companies are hiring to add senior firepower, not to train up the next cohort. If you're early-career, that ratio is the hard truth of this snapshot, and it's why the junior search takes longer than the postings-count alone would suggest.

Which companies are hiring fastest?

Weloglobal, by a wide margin, with +104 net roles in 28 days and 104 active — its whole board turned over into new demand this month. After that the pace drops sharply:

• Weloglobal — +104 net (104 active)
• Mistral AI — +42 net (42 active)
• Neura Robotics — +38 net (38 active)
• Cerebras — +29 net (29 active)
• Palo IT — +27 net (27 active)
• Innodata — +22 net (22 active)
• SpaceX — +16 net (28 active)
• Doctolib — +15 net (17 active)
• SentinelOne — +14 net (18 active)
• Grafana Labs — +14 net (22 active)

A few patterns are worth calling out. Most of these companies have net-new numbers almost equal to their active count — Mistral AI at +42/42, Neura Robotics at +38/38, Cerebras at +29/29 — which means these are fresh hiring pushes, not slow accumulation. When net and active match, the whole board went up this month.

SpaceX is the interesting exception: +16 net but 28 active, so it's been carrying AI/ML openings for a while and topped them up. It's also the reminder that "AI/ML jobs" isn't only labs and infrastructure startups — it's aerospace, healthcare (Doctolib), security (SentinelOne), and observability (Grafana Labs) all needing the same skill set. The demand is diffusing into companies whose core product isn't AI.

The frontier-lab and hardware names you'd expect show up too. Mistral AI at +42 and Cerebras at +29 are the model-and-silicon end of the spectrum, and Neura Robotics at +38 points to robotics pulling ML talent hard — a corner of the market that's been quietly aggressive. Note that several of these are European (Mistral, Neura, Doctolib, Palo IT), which fits with only about a third of the overall market being remote and a lot of hiring being tied to specific locations.

How much of this is remote?

34% of open AI/ML roles are remote. That leaves roughly two-thirds tied to a location, which is a meaningful constraint if you're not willing or able to relocate. Given how many of this month's fastest-hiring companies are European, and given aerospace and hardware employers like SpaceX and Cerebras that need people on-site by nature of the work, the remote share isn't surprising — you don't build robots or chips over Slack. If you're remote-only, you're effectively fishing in a third of the pond, and the senior skew means the roles that survive as remote tend to want more experience.

Is this a good time to look?

It's a good time if you're mid-to-senior and an engineer, and a harder one if you're junior or non-engineering. The market is net-positive, 569 brand-new employers entered in the last month, and the disclosed bands sit comfortably in the $150K–$230K range for the core titles. That's a real, funded, hiring market — not a frozen one.

But the composition matters more than the headline. Nearly half of specified roles are senior or above, principal openings outnumber junior almost four to one, and 95%+ of everything is engineering. The doors that are open are mostly open for people who can build production ML systems and point to having done it before. If that's you, apply broadly — the churn (3,659 opened in 28 days) means new roles surface constantly, so checking weekly beats a one-time sweep.

The takeaway for job seekers

If you're a mid-to-senior ML or AI engineer, go where the net-new hiring actually is this month rather than only chasing the famous names: Weloglobal alone opened 104 roles, and Mistral AI, Neura Robotics and Cerebras each stood up a full fresh board. Those are the places most likely to have unfilled slots and fast-moving pipelines right now. Because only 18% of postings show pay, use the disclosed $159K floor for ML/AI Engineer as your anchor and ask for a range early — most employers here won't volunteer one. If you're early-career, be realistic about the 4:1 principal-to-junior ratio: target the companies posting genuine junior roles specifically rather than applying into senior reqs hoping to be leveled down, and treat non-remote ro