Самые популярные языки программирования в 2026 году – и несколько хороших новостей о количестве вакансий разработчиков

• Python лидирует, но старые языки по-прежнему сохраняют свою популярность.
• Искусственный интеллект может создавать приложения, но языковые рейтинги по-прежнему рассказывают историю.
• Данные о вакансиях дают неожиданный прогноз на будущее программирования.

Несколько дней назад мой редактор отправил мне сообщение в чат. Она спросила: "Помните, как хорошо публиковались ваши статьи о самых популярных языках программирования? Как это изменилось с появлением vibe coding? Важны ли эти рейтинги сейчас?"

С этими словами я приготовил первую из множества чашек Pacific Bold и приступил к работе. То, что я обнаружил, оказалось совсем не тем, что я ожидал.

Итог?Python, Java и JavaScript по—прежнему лидируют в чартах, но произошли и другие заметные сдвиги - как с прошлого года, так и за те 10 лет, что я отслеживаю популярность языков программирования. Но я также обнаружил, что с тех пор, как появились инструменты для программирования agentic vibe, количество вакансий разработчиков из месяца в месяц увеличивалось. Я могу это доказать, и приведу это ниже с помощью графиков.

В последний раз я изучал язык программирования в 2025 году. Это было во времена создания генеративного ИИ. Примечательно, однако, что она была опубликована до появления таких программистов, как Claude Code и ChatGPT Codex.

В этой статье мы немного усовершенствовали нашу игру в языковой анализ, создав собственный индекс популярности языка программирования.

Проблема определения популярности (как известно многим политологам) заключается в том, что все зависит от того, кого вы спрашиваете, когда вы спрашиваете и как вы спрашиваете. То же самое относится и к языкам программирования. Оцениваете ли вы репозитории на GitHub и их основные языки? Как насчет объявлений о вакансиях на определенных форумах? Или, возможно, вы оцениваете количество сайтов, найденных в поиске Google?

Ни один из них сам по себе не дает четкой картины состояния языков программирования. Вместо этого я использовал девять отдельных опросов, присвоил вес позиции каждого языка в каждом опросе, а затем объединил эти значения в единую шкалу.

Мой индекс за 2025 год включал данные за период с 2023 по 2025 год. Но поскольку программисты стали разрушительной силой в основном в 2026 году, я ограничил индекс этого года опросами середины или конца 2026 года. Это позволило мне увидеть влияние агентурного кодирования, отраженное в самих данных.

• PYPL pulls data from Google search activity (October 2026).
• Tiobe aggregates mentions and listings across popular websites (October 2026).
• GitHub publishes most-used languages by GitHub developers (October 2026).
• Stack Overflow publishes survey results with top-used languages (October 2026).
• The Consensus blog indexes data from "builders of major software infrastructure systems" (October 2026).
• Tech Talent Portugal indexes survey data from Portuguese developers (May 2026).
• GigaZine indexes survey data from Japanese language developers (August 2026).
• Oxylabs job listings analysis (May 2026). We covered Oxylabs last year.
• JetBrains language usage analysis (August 2026).

While the top entries are fairly consistent, our survey aggregate table shows that the results overall are quite chaotic. They were last time, too.

After processing the data using the ZDNET framework, I selected the top 20 programming languages across all nine data sources. As you can see, Python, Java, and JavaScript still rule.

I found a few things quite interesting. Note that both Visual Basic and VBA are on the list. In the survey aggregate, Fortran and COBOL were also on the list. While clearly languages from our deep, dark past, the presence of these Jurassic languages indicates that there is a definite need for ongoing support of legacy systems.

I’ve actually done these language popularity studies going back ten years, although the earlier ones didn’t have the sophisticated multi-study aggregation methodology I’ve been using recently. Unfortunately, the earliest articles are inaccessible now, having been replaced with updated data years ago. Even so, I was able to find some of my original 2016 usage data in my server’s media database. I was able to incorporate this decade-old data into the following language preference migration chart.

While there has certainly been a lot of movement over the years, most languages remain relevant, even if their ranking has increased or decreased over the decade.

The movement of Python into the top rank, and the decline of the more traditional C-like languages of Java and C++, indicate eager adoption of Python’s ease of use and capability, as well as its overwhelming level of adoption in AI workloads.

Also: ‘I can’t stop’: 80% of developers find AI coding more addictive than helpful

While Python is a hybrid tool, which is compiled into byte codes and then interpreted, the more traditional compiled languages have all dropped in usage. This reflects not only the productivity benefits of languages like Python, but also the rise in script-level languages for web use, and the fact that processors are so much faster that pure compilers don’t offer the same performance benefits they did in years past.

As for the question about how rankings have shifted with vibe coding, as we can see, a full 30% (and all three top languages) didn’t move at all. But the big question — whether language rankings even matter anymore — is coming up next.

Even a couple of years ago, if you were going to write code, you would have had to do it like a caveperson. You had to write every single line of code by yourself, maybe with some helpful advice from an AI. Now, you can go all the way from nothing to commercial apps sold on the Mac App Store, without writing a single line of actual code.

While there has been considerable coverage of the recent wave of layoffs (Amazon, Meta, Oracle, Xbox), there’s also a strong case to be made that the idea of a developer apocalypse is overblown.

When I started working on this article, I wanted to involve two types of data: language-by-language usage information, and developer growth or decline rates.

While it’s very easy to find articles about developer layoffs, it’s a bit more challenging to find month-by-month data on available developer jobs. That said, I found a very interesting collection of datasets, buried fairly deeply on the website of the Federal Reserve Bank of St. Louis. The St. Louis Fed (part of the United States’ central banking system) is well known for maintaining the Federal Reserve Economic Data library, affectionately known as FRED.

Also: 75% of developers I surveyed prefer Claude Code over Codex – here’s why

FRED has two interesting datasets from the job posting site Indeed. The first one is Software Development Job Postings on Indeed in the United States, with daily index offsets going all the way back to February of 2020. The second one is overall Job Postings on Indeed in the United States, also with daily index offsets going back to February 2020.

Before I show you my first chart, I need you to wrap your head around what, exactly, this is indexing. Rather than an absolute number of jobs, the index offset shows the change in job counts from an arbitrary index date, in this case Feb. 1, 2020 (which just so happens to be right before the pandemic hit).

What all that means is that the following chart shows the change in job listings available, but they’re not to scale. This is important, because it looks like the number of developer jobs in 2022 exceeded the number of all jobs. That’s not true. Instead, the growth of developer jobs exceeded the growth of all jobs as compared to the index date of Feb. 1, 2020.

As you can see, developer jobs took a big jump in the years following the pandemic, as infrastructure build-outs were rapidly updated to accommodate remote work and other urgent changes in IT requirements. But once the pandemic faded, the surge dropped.

We’re still seeing that in terms of layoffs. While experts might attribute some layoffs to AI, another reason people are being let go is that companies overhired at the beginning of the decade. If you follow the dotted line as the index point, you can see that programming job growth still hasn’t recovered from the pre-pandemic index point.

Also: Claude Code vs. Codex: Why I use both and how to choose one

But if we can look at job growth relative to an arbitrary point in time, we’re not limited to February 2020. We can refactor the index offset data to base growth analysis from any other starting date.

So, I refactored, basing my starting point on Oct. 1, 2025, a year ago. October 2025 is, not coincidentally, when OpenAI’s Codex was first released for general availability. I tried it a month earlier, before it was fully in general availability. I could immediately see the potential.

Claude Code had been out for a while. But the second big milestone I wanted to showcase was when Claude Code’s auto-mode went live. This is a permissions level in Claude that lets it run for a lot longer, without human supervision, and yet still be fairly safe. I found auto-mode to be game changing in some of my development projects.

Look at that chart. What do you notice?

I noticed that growth in developer jobs is increasing faster than overall job growth. In fact, even as overall job growth declined from February to June 2026, demand for coders increased in that same time period. In fact, as I mentioned before, since the advent of agentic vibe coding tools, month-over-month growth in developer jobs has increased!

Also: How I made a paid Mac app in 11 days with Claude – without writing a single line of code

At the beginning of this article, I promised you something unexpected. Given how much developer attention is required in wrangling AI coding tools, I didn’t exactly expect job growth to plummet. But the one thing I really did not expect was to see developer job growth outperform the overall job market, especially now that AI agentic coding tools do such a good job of building working code.

Now that it’s possible to get so many more projects done in a given block of time using these amazing tools, it seems like companies are hiring more people to keep up with the opportunities offered.

What do you think? Have you seen more or fewer developer job opportunities since agentic coding became a thing? Let us know in the comments below.

You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.