Count the open tech roles with "Python" in the title. On October 28, 2026 we found 51.
Now count the roles that name Python anywhere in what they ask for. That is 5,637 of the 16,055 open tech roles we counted the same day, at 540 of the 776 companies with anything open. More than a third of everything, and any search that only reads job titles would show you less than 1% of it. (These are distinct roles, so one role posted in ten cities counts once.)
So the first thing to know about Python jobs in 2026 is that they almost never say so. The title says "Software Engineer" or "Research Engineer" or "Solutions Architect", and the Python is in the body. Which raises the obvious next question: what are all these jobs, then?
Mostly not data science
We sorted the Python roles by what their titles mention (the first match wins, so a "Machine Learning Infrastructure Engineer" counts as ML/AI):
| What the title says | Roles |
|---|---|
| General software or backend engineering | 2,041 |
| Machine learning or AI | 1,010 |
| Infrastructure, platform or SRE | 598 |
| Solutions, sales or customer engineering | 518 |
| Security | 367 |
| Data science or analytics | 344 |
| Data engineering | 208 |
| Forward deployed engineering | 188 |
| Something else | 363 |
Data science and data engineering together come to 552 roles, under 10% of the total. If your picture of a Python job is a notebook full of pandas, the board disagrees (politely, with a lot of backend services).
Two rows surprised us. 1,010 roles are ML or AI, which is about what you would expect. But 518 are customer-facing (solutions architects, sales engineers, customer engineers), and another 188 are forward deployed engineers. Our read: Python is a customer-facing skill now, not only the language you ship in.
Who is hiring
Anduril Industries has the most, 530 roles naming Python, about 9% of the total. After that:
| Company | Python roles |
|---|---|
| Databricks | 233 |
| OpenAI | 205 |
| Anthropic | 161 |
| Cloudflare | 95 |
| Snowflake | 94 |
| Celonis | 88 |
| Canonical | 87 |
| Nebius | 87 |
| Affirm | 76 |
| Mistral AI | 73 |
| Airwallex | 73 |
The top twelve hold about a third of all Python roles, and 247 of the 540 companies have only one to three open. Compare that with the systems languages earlier this month: twelve employers held 57% of Rust roles, and Anduril alone held a quarter of C++ roles. Python is the opposite. It shows up at 540 of the 776 companies with open roles, about 70% of them, which is good news if you do not want your job search to depend on one defence contractor's headcount plan.
What Python travels with
2,090 Python roles (37%) name no other programming language (out of the 28 we check for). At the other end, 1,025 name four or more.
The rest of the stack looks like this. AWS shows up in 1,702 of them, SQL in 1,482, Kubernetes in 1,458. And then the one we did not expect: Go, in 1,370. Leave SQL aside (it is a query language, and 1,482 of these roles list it) and Go is the programming language that shows up next to Python most often, ahead of TypeScript (947), Java (942), C++ (820) and Rust (546).
We think that is the backend story again. A team that writes its services in Go and its tooling, pipelines or ML in Python is a very normal team in 2026, and it would make sense for that team to want people who can read both.
Where we can separate the core of a job from the things it mentions in passing, Python is core in 2,665 of 3,702 roles, about 72%. So where we can tell, when Python is there it is usually the point.
Pay, and getting in
3,333 Python roles publish a yearly salary we can express in US dollars (converted where the post used another currency). Their median floor is $165,000, and the median ceiling, from the 3,313 that give one, is $226,550.
That is almost exactly the middle of the board: the median floor across all 9,473 roles with a yearly salary is $166,000. Python does not carry a premium the way the systems languages did when we counted them earlier this month (Rust and C++ both sat around $190,000). It is the baseline, which makes sense for the language a third of the market uses.
The upside is the door. 402 Python roles are junior, 7.1%, against 5.8% across all open tech roles (and 2.4% for Rust when we counted it). If you are early in your career, Python's door is wider than the market's.
On remote work: 123 roles (2.2%) are labelled remote with no location restriction. 1,704 are remote within a country or region, 2,189 are hybrid and 1,621 are onsite. Remote Python jobs exist, but nearly all of them come with a map attached.
If Python is your language
Search by stack, not by title. Fifty-one titles against 5,637 roles is the whole argument (and it is why roles on remoet.dev/jobs show their stack).
Learn to read Go. A quarter of these roles name it.
Take the cloud seriously. AWS (1,702), Kubernetes (1,458) and Terraform (789) all show up more often than PyTorch (407), the most common Python library in these roles.
Look past engineering titles. Solutions and forward deployed roles add up to 706, one in eight Python roles, and talking to customers is the job.
If you star the companies you care about, their new roles come to you as they post, which beats running that title search every morning.
Where this could be wrong
We match "Python" in the stack a model extracts from each post, so "Python or Go" and "Python and Go" look the same, and a passing mention counts like a requirement (the 72% core figure is our best check on that). The job-type table sorts titles with simple patterns, first match wins, and "Something else" is a real 363 roles we could not place. Salaries cover the 59% of roles with a yearly figure we could convert to dollars. And it is all one day, October 28, 2026.
If you write Python for a living and your job looks nothing like this, we would like to know what we are missing. Come tell us in Discord.
