"49,000+" makes this the least responses the survey has gotten since 2016 ("over fifty thousand"), every year in between has been in the 65-100k range. Seems as though enthusiasm around SO has diminished significantly over the past year.
My guess is that pro-AI devs have abandoned the site, and anti-AI devs are upset with their collaboration with AI companies.
For some context on demographics of this survey: I signed up for SO in about ~2011 and haven’t participated in SO since about 2015. But I have generally positive sentiment towards the survey so I do it every year.
A lot of comments in this thread make the assumption that the only respondents are people who actively use SO but that isn’t true. I just get notified about the survey through email every year and respond.
Yeah those most likely to respond are those who find SO to be important and value it for whatever reason. And inversely those unlikely to respond are those who dislike SO. So the nature of SO itself will be reflected in the demographics of the survey responders.
C# as popular and commonly used as java? Hmm either I'm woefully behind the times, most java devs aren't answering SO surveys, or the data is wrong. Then again the fact that C# is used in indie game dev probably gives it a serious boost in the 'evangelical user' category. It is a nice language, I just find java to be far more used in industry.
Also, the editors? I'm sorry I've simply never gotten vsCode when jetbrains and neovim exist, much less N++.
Regardless I think I have to acknowledge that maybe I'm not your average dev. TBH your average dev is probably very happy coding up react widgets in vsCode, and I'm the grouchy greybeard behind the times.
Probably not as popular, but C# is absolutely up there, even though it doesn't get a lot of mindshare. I work for a small firm that does consulting for enterprise customers and I see a lot of C# around.
Europe especially is basically MS's backyard.
In Australia, C# feels more popular than Java. I used to take on freelance projects in both and had far more work in C#.
I agree it'll pop up more with "evangelical users" though, 100%.
I've often thought that when you see Mac and Linux more strongly represented than you'd expect, I do wonder when people dual boot Windows 95% and Linux 5%, are they ticking the box for Linux? Does the VB.net hostage even bother to fill in the survey?
Agree on VS Code, it's OK if you can't find support elsewhere, but the JetBrains stuff is just in a different class.
My beard is only partially grey, but I agree with you.
On C# this could be your country/sample. As far as I remember, though this is from almost a decade ago now, C# is more popular than Java in the UK and Australia but Java comes out ahead in the US.
Recall that this is a StackOverlow survey, and reflects their history and audience. [0] Even still, I'm also surprised.
[0] https://www.joelonsoftware.com/2006/09/01/language-wars/
I believe Java is more popular for enterprise and web apps.
.NET is widely used for videogames (games made with Unity, Godot, Unigine engines, also internal tools and game servers), desktop software, embedded software. Java is rarely used for these things, hence the results of that survey.
I think you're grossly overlooking swathes of relatively small, non-tech companies that just treat Microsoft as the default for everything. Need a programming language? C#'s there. Need a coding text editor? VSCode.
Vscode is free, which is a huge positive for it. For lots of places it can be hard to get budget for dev tools. A lot of it can depend on your industry or region
Stack overflow is built using C# and I remember there being a large Microsoft stack developer community on SO back when I used it. That might skew the results
In Norway, C# seems to have surpassed Java in the finance sector several years ago. C# is also much bigger than Java in game development.
You seem so confident, do you have insights into the entire industry?
>Regardless I think I have to acknowledge that maybe I'm not your average dev. TBH your average dev is probably very happy coding up react widgets in vsCode, and I'm the grouchy greybeard behind the times.
You should work on your reading comprehension
I usually participate, this year I don’t remember being notified about it, but I guess that probably goes to show I haven’t been using SO nearly as much as I used to due to AI tools.
On a separate note, why are things like cargo/apt/homebrew being put into the same category as Docker and Terraform, which also includes entire cloud platforms themselves (Azure, GCP,etc…). Might as well throw programming languages in there for good measure xD
How is it that Windows always dominate these Stack Overflow developer surveys, but people talk like Mac is the one that's ubiquitous among devs?
And I hate these big accept cookies pop ups or whatever they are called that impede use of a website unless you accept. I couldn't find an option to reject all cookies.
Mac users talk like Mac is ubiquitous among whatever niche they're in. That's a trope as old as the Year of the Linux desktop.
I remember my "Linux friend" sent me an excited email when WSL was announced, signalling that somehow this meant that Microsoft has conceded. My response: "Finally the year of Linux on desktop". I think that was his last email ;-)
The most interesting thing to me is that some people actually prefer the WSL2 experience over using Linux, not just because of the Windows desktop, but because of the WSL2 experience itself.
I think Distrobox gives you the same basic experience but it's even a bit nicer as it's even lower overhead and doesn't have the weird memory management issues that WSL2 has. But this is really interesting to me; Microsoft clearly struck a nerve with the WSL2 workflow. It's more than just a workaround for the fact that some stuff doesn't run or doesn't run well on Windows, which is counterintuitive.
You can also get the same basic experience on macOS using Orbstack, which I highly recommend in general. (Though there are alternatives that are not subscriptions, but I only use macOS under duress at work, so I don't personally pay for it.)
Stack Overflow is a bubble, but Hacker News and other dev-focused online places are probably an even larger bubble. There's a lot of software developed that isn't particularly visible, and where the developers are not as active online as the ones here.
The "sexy" development stuff (FOSS web frameworks, LLMs, etc.) have a habit of happening on Macs.
Not all of it, mind.
> "sexy" ... web frameworks, LLMs ... on Macs
Oh my, I don't even know where to start on this one. If programming is sex, web programming is submission. Sure it sounds fun, but not for regular people. As for LLMs I can't imagine a situation where I'd want to run them on Macs and ditch my Nvidia graphics.
The real reason is that Windows users are just not very loud.
>The real reason is that Windows users are just not very loud.
And why's that?
I dunno. I work on my PC. At some point I feel like 64GB RAM is not enough. I open Amazon and order another 64GB. Some other time I think if I upgrade my CPU my IDE experience would be better. Depending on the generation, I either order another CPU or get a whole new PC altogether (these days I also redistribute my GPUs between the old and the new one).
And so on. I assume everyone else with PCs does the same. So when I go to web forums I don't usually talk about that.
Total for Linux when distros combined dominates Windows
It's simple and observed over a number of years. HN is an echo chamber. Mac users drown out and shout down other users here. Other users go elsewhere for discourse. As a case in point look at the other replies to your comment.
On a separate note it's sad to see that Ubuntu has fallen behind.
uBlock Origin has a Cookie Notices filter which blocks those. NoScript worked for this one, too.
For some reason excited young devs learning javascript are convinced that they need a mac
> Stack Overflow is a frequent destination for information. A strong majority (82%) visit at least a few times per month, with 25% visiting daily or more often.
A survey among visitors of a tobacco shop shows that 82% of people smoke frequently.
The numbers don't add up to 100%. I personally draw strong conclusions about the overall quality of statistics when I find an obvious error just from scanning over it.
Highly trust 3.1%
Somewhat trust 29.6%
Somewhat distrust 26.1%
Highly distrust 19.6%
========================
sum 78.4%
source: https://survey.stackoverflow.co/2025/ai#2-accuracy-of-ai-too...
No Emacs entry under editors, and no Clojure in the languages section? Somebody on the survey team had a bone to pick.
They're in the write-ins for their respective categories. Emacs with 0.1%. Clojure with an impressive 0%!
I thought Emacs was an operating system :-)
No surprise seeing as they authored this: https://stackoverflow.blog/2020/11/09/modern-ide-vs-vim-emac...
Maybe they think the certain generation who uses Emacs has died off or something.
Last years results had Emacs at 4.2% https://survey.stackoverflow.co/2024/technology . I would expect it to be roughly the same. Emacs seems to hover around this usage percentage.
I always jump straight to the IDE section because that's the most intersting to me https://survey.stackoverflow.co/2025/technology#1-dev-id-es
A few notes:
- VS Code at #1 (of course) with 75%, Visual Studio at #2 with 29%
- Vim at 24% and Neovim at 14%, which seems pretty whopping to me (I wonder how much overlap is there, the survey is clearly "click all that apply")
- Cursor is 17.9%, wild
- Nano is 12.5%, lol
- Sublime Text is 10.5%, impressive given how much the IDE market has changed
- Zed is 7.3% I wonder if that's enough to be Ramen profitable for a small team? (Not familiar with what Zed is charging for these days)
And Emacs was banished to a write-in this year, when it was near the middle of the pack at 4.2% (Emacs) and 0.4% (Spacemacs) last year.
https://survey.stackoverflow.co/2024/technology#1-integrated...
It's also been fairly stable over the last few years, so it's hard for me to believe that it just suddenly dropped off to 0.2%:
| 2021 | 5.30% |
| 2022 | 4.51% |
| 2023 | 4.69% |
| 2024 | 4.2% |Vim is very likely to be an editor that complements a full IDE if you're just doing something quick. If you asked me I'd also check Vim even though I'm usually on Jetbrains.
Yeah that makes sense, here I'm just curious if 38% of developers really use Vim (i.e., Vim [24%] + Neovim [14%]), which seems really high to me (even if it's assumed a lot of these are using Vim as a second editor). Or if it's just that a lot of folks are checking both Vim and Neovim.
Zed lets you pay for AI, and I’m a happy customer. Their agent feature works well with Claude models IMO and the full integration and editor experience is excellent.
Can anyone make sense of the sankey chart under "Developers at all levels are exploring the evolving AI landscape through Stack Overflow"? How does "Large Language Model" flow into "Tailwind CSS 4"?
I think it was made by an AI itself
I joke, but I think it’s depicting relative proportions of how many people who were “interested in” Gemini are now “interested in” other things and how different subgroups’ “interested in” thing changed over time. So there was a sizable group that became interested in Tailwind at one point
As in, Gemini users are interested in Tailwind? Is this suggesting that Gemini is suggesting tailwind, and people are using it? It's very weird
I didn’t realize you can click into each thing to get more details. The question asked for this one was:
“Which rising technology of the added as new tag or tag subject area on Stack Overflow in the last three years have you used regularly in the past year, and which do you want to work with over the next year? Please check all that apply.”
The chart is displaying every pair of (worked with => want to work with) that had at least 500 respondents. So it’s not really a sankey diagram, each ribbon just signifies a self-contained pair. So yeah many people who worked with Gemini indicated that they want to work with Tailwind (among other things). It’s not suggesting that Gemini itself recommended those technologies.
It is weird. The biggest thing for me is that when you click through to the details you can filter to those who indicated that they are learning to code, and and Gemini had by far the largest “worked with” amount for that group. Which is interesting — suggesting the next generation of coders are going to have all learned with the “help” of AI (yikes)
Did anyone observe that .NET core is far ahead of Spring in popularity and beats other backend frameworks in other measures as well?
They got rid of the salary by languages used category?
I liked seeing the languages that "paid more".
Some comparisons:
Job satisfaction (happy/complacent/unhappy):
- 2025: https://survey.stackoverflow.co/2025/work#5-job-satisfaction (24.5/47.1/28.4)
- 2024: https://survey.stackoverflow.co/2024/professional-developers... (20.2/47.7/32.1)
- ...
- 2015: https://survey.stackoverflow.co/2015#work-satisfaction (36/51.4/11.6, complacent includes "somewhat" and "so-so")
Work environment (remote/hybrid/in-person):
- 2025: https://survey.stackoverflow.co/2025/work#1-work-environment (45/37.1/17.9, but I counted "your choice" as remote)
- 2024: https://survey.stackoverflow.co/2024/work#1-work-environment (38/42/20)
- 2023: https://survey.stackoverflow.co/2023/#section-employment-wor... (41.41/42.18/16.41)
- 2022: https://survey.stackoverflow.co/2022/#section-employment-wor... (42.98/42.44/14.58)
- ...
- 2017: https://survey.stackoverflow.co/2017#remote-work (11.1/52.8/31.8, hybrid includes all answers of "sometimes", the 4.2 who said "it's complicated" not included)
Salary (engineering manager/full-stack/academic, not adjusted for inflation):
- 2025: https://survey.stackoverflow.co/2025/work#salary (130k/73k/57k)
- 2024: https://survey.stackoverflow.co/2024/work#salary (124k/63k/49k)
- 2023: https://survey.stackoverflow.co/2023/#section-salary-salary-... (124k/71k/54k)
- 2022: https://survey.stackoverflow.co/2022/#section-salary-salary-... (112k/66k/55k)
- 2021: https://survey.stackoverflow.co/2021#section-salary-salary-b... (96k/56k/49k)
- 2020: https://survey.stackoverflow.co/2020#work-salary-by-develope... (92k/54k/41k)
- 2019: https://survey.stackoverflow.co/2019#work-_-salary-by-develo... (95k/57k/38k)
Employment status (employed/contractor/unemployed, not counting students/retired/unknown):
- 2025: https://survey.stackoverflow.co/2025/work#1-employment-statu... (69.8/13.9/4.6, all are less for some reason)
- 2024: https://survey.stackoverflow.co/2024/work#1-employment-statu... (75.3/16.4/7.8)
- 2023: https://survey.stackoverflow.co/2023/#section-employment-emp... (75/15.91/6.32)
- 2022: https://survey.stackoverflow.co/2022/#section-employment-emp... (74.4/14.95/6.44)
- 2021: https://survey.stackoverflow.co/2021#section-employment-empl... (67.26/9.65/5.02, like 2025)
- 2020: https://survey.stackoverflow.co/2020#work-employment-status (74.4/8.9/4.2)
- 2019: https://survey.stackoverflow.co/2019#work-_-employment-statu... (79/9.8/10.8)
Did the people satisfied with their jobs increase because the unsatisfied ones are no longer employed?
Why is Cargo the top cloud infrastructure tool? I thought it was just Rust’s package manager?
Pretty cool to see Redis as the most used Database for AI Agents!
Why?
Meh, as usual this survey gives completely skewed view of the industry by being dominated by students and junior devs.
Only ~14% of respondents had <=5 years of dev experience. And 55% were >10 years dev experience.
One thing I did find funny though was this survey found that devs overwhelmingly visit stackoverflow aka the site that puts the survey out found that so many people that use the site, use the site.
> And 55% were >10 years dev experience
This doesn't mean they aren't Juniors though
It does. Whatever they are, they aren't junior developers any more. As a hiring manager, I wouldn't hire a dev with >10 years experience into a junior role - if they're still at junior level, it's not that they're a junior, it's that they're unable to progress, which is a whole different issue.
I think the person you're responding to has likely read this quote or one like it, and perhaps you haven't? I assume that they're referencing it in an offhand way.
https://news.ycombinator.com/item?id=4627373
> Some people have 10 years of experience. Others have 1 year of experience 10 times.
I found that odd.
Since I started using ChatGPT, I rarely visit stack. If I need some code ChatGPT will tailor it to my example and I can get up and running quickly to figure out how it works.
That's <5 years of experience in coding. I suspect most people say they've been programming since their teens
24.8% of respondents have been programming for less than 5 years
Sounds like Stack Overflow is your dad's programming help resource. The kids are somewhere else.
it would be rather odd if 60% of the respondents didn't use the site, wouldn't it?
GitHub Copilot and Cline are significantly less condescending.
If SO disappeared, I wonder what effect it would have in the training of LLMs as languages and tooling continue to evolve.
I feel like the main corpus for GHCP, at least, is probably just GitHub.
SO is filled with all sorts of questions from 10-15 years ago that aren't up-to-date with the languages and tooling of today. If you develop in a language that is substantially different from what it used to be (JavaScript, Python, etc.) that is problematic.
The site got to be a pain to try to participate in long before modern AI came around.
This is the first year I didn't care to participate, but I've been away from SO for about 4 years now.
I can repeat the same question to AI all day and it is happy to help me.
Which is something you shouldn't ever do to a human. SO made you search and think before posting, while AI trains people to outsource thinking.
For all its failings, SO strived for a very high signal-to-noise ratio, and it achieved it.
As others have said this before, AI is a tool and it is not supposed to nor should replace thinking. Unfortunately many people does not abide by this. See the other submission where people made a 128k line PR to a project: https://news.ycombinator.com/item?id=44729461.
It is clear that the guy responsible for this PR outsourced thinking. I would call it "misuse of LLMs". LLMs get a bad rep because of individuals who completely outsource thinking.
Some added commentary on the concentration of programming languages: https://open.substack.com/pub/therosen/p/analyzing-the-stack...
I think people are abandoning SO because AI often gives better answers. And also SO search results on Google just got so much worse a few years ago, you search Google for "how to do X in framework Y" and get SO results about doing X in framework Z.