js2025.yml

locale: ua-UA
translations:
  ###########################################################################
  # General
  ###########################################################################

  - key: finish.share_message.js2025
    t: The State of JS 2025 survey is now open! {link}

  - key: general.results.description
    t: Щорічне опитування про останні тенденції в екосистемі JavaScript за 2025 рік.

  ###########################################################################
  # Survey Intro
  ###########################################################################

  - key: general.js2025.survey_intro
    t: >
      JavaScript, you OK? You've barely created any new front-end frameworks this year.


      After a decade of rapid iteration, it seems clear that the ecosystem has stabilized over the last few years. For example, did you know that “newcomer” Svelte is *9 years old*? Which, in JavaScript framework years, is basically ancient!


      Instead, the main battle has moved to the realm of meta-frameworks, with Astro making a serious attempt at Next.js's crown – while over in build-tool land, it seems only a matter of time until Vite overtakes webpack. 


      Or maybe something else entirely will appear out of the Rust ecosystem and suddenly take over, bringing a premature end to our well-deserved era of peace and quiet.  


      Whatever happens, one thing's for sure: you can help us figure it all out by taking this year's State of JavaScript survey!

  - key: introduction.js2025
    t: >
      Коли нещодавно вийшла довгоочікувана Nintendo Switch 2, минуло вже вісім років від появи оригінальної Switch.


      Чи можете ви уявити, як це було для тих бідних розробників ігор? Вісім довгих років розробки для тієї самої платформи — без нових API, без нових можливостей, без нових багів, без нових версій, що ламають вашу збірку… Бідолашні!


      На щастя, про це *нам* ніколи не доведеться хвилюватися, адже боги JavaScript і далі дбають про те, щоб кожен новий рік приносив свою порцію пробл… тобто *інновацій*!


      Тож приєднуйтесь до мене ще раз, щоб переглянути результати цього опитування й озирнутися на засіки 2025 року.


      <span class="conclusion__byline">– Sacha Greif</span>

  ###########################################################################
  # Survey FAQ
  ###########################################################################

  - key: faq.how_long_will_survey_take_js2025
    t: Скільки часу займе опитування?
  - key: faq.how_long_will_survey_take_js2025.description
    t: >
      Залежно від того, на скільки питань ви відповісте (всі питання не обов'язкові), 
      заповнення анкети займе орієнтовно 15-20 хвилин.

  - key: faq.survey_design_js2025
    t: Як було розроблено це опитування?
  - key: faq.survey_design_js2025.description
    t: >
      Опитування було розроблено унаслідок [відкритого процесу розробки](https://github.com/Devographics/surveys/issues/307) із залученням постачальників браузерів і спільноти веброзробки.

  - key: faq.results_released_js2025
    t: Коли будуть опубліковані результати?
  - key: faq.results_released_js2025.description
    t: Опитування триватиме із 1 жовтня до 1 листопада 2025 року, і його результати будуть опубліковані незабаром після завершення.

  ###########################################################################
  # Overview
  ###########################################################################

  - key: overview_1_frontend_fatigue.js2025
    t: |
      Клієнтські фреймворки

  - key: overview_2_whats_next.js2025
    t: |
      Метафреймворки

  - key: overview_3_back_in_business.js2025
    t: |
      Серверні фреймворки

  - key: overview_4_build_it.js2025
    t: |
      Збирачі коду

  - key: overview_5_static_typing.js2025
    t: |
      Статична типізація

  - key: overview_6_other_languages.js2025
    t: |
      Інші мови

  - key: overview_7_conclusion.js2025
    t: |
      ### Відкрито зараз
      ## Стан React 2025

  ###########################################################################
  # Takeaways
  ###########################################################################

  - key: features.js_new_features.takeaway.js2025
    t: >
      Хоч **Temporal** усе ще очолює рейтинг, рівень зацікавлення знизився за рік, що цілком логічно, адже він почав перехід від теоретичної пропозиції до реалізації у Firefox.

  - key: features.top_currently_missing_from_js.takeaway.js2025
    t: >
      У топ-5 немає змін, але варто зазначити, що інтерес до **Signals** трохи знизився, як порівняти з минулим роком.

  - key: features.native_types_proposal.takeaway.js2025
    t: >
      Хоч 32% респондентів і сподіваються на **динамічні типи** у JavaScript, та якщо офіційне підтримання типів колись з’явиться, то набагато ймовірніше як **анотації подібно до TypeScript**, які посіли перше місце.

  - key: features.js_pace_of_change.takeaway.js2025
    t: >
      Метою цього питання було приблизно уявити, наскільки добре респонденти встигають за новими функціями, і результати показують, що більшості з нас складно навіть дізнаватися про них — навіть не пробувати. Сподіваємося, результати цього опитування допоможуть із цим!

  - key: features.reading_list.takeaway.js2025
    t: >
      **Список читання** дає респондентам змогу дізнаватися більше про конкретні пункти після проходження опитування, і цього року, крім ручного додавання, почало функціонувати автоматичне додавання на основі відповідей. 

      Ця зміна призвела до того, що список став значно більш орієнтованим на бібліотеки, як порівняти з попередніми роками, коли в ньому було більше функцій мови.

  - key: libraries.all_tools_experience.takeaway.js2025
    t: >
      **Vite** уже кілька років є найбільш улюбленою бібліотекою, але тепер вона обійшла **React** і стала також другою за рівнем використання, і, схоже, готова посісти перше місце вже наступного року. 


      Водночас **Next.js** є 13-м найулюбленішим і 5-тим найменш улюбленим проєктом, а ще зібрав найбільше коментарів загалом — що демонструє, наскільки поляризованим може бути ставлення до фреймворку.

  - key: libraries.tools_arrows.takeaway.js2025
    t: >
      JavaScript-бібліотеки зазвичай демонструють "бумеранг-патерн" із часом: на початковому етапі позитивні відгуки (лінія рухається вправо) сприяють росту використання (лінія стрімко підіймається), але користувачі стикаються із що раз неприємнішими крайніми випадками та проблемами під час реального впровадження, через що думки стають негативнішими (лінія повертає вліво). 


      Це означає, що будь-який проєкт, який рухається вправо (як-от **Vue.js**) або навіть просто залишається вертикальним (як-от **Astro**), заслуговує на увагу!

  - key: libraries.tier_list.takeaway.js2025
    t: >
      Завдяки рівню вдоволеності понад 90%, **Hono** і **Bun** цього року цілком заслужено потрапляють до S-tier.

  - key: libraries.tools_ratios_over_time.takeaway.js2025
    t: >
      Нагороду за найбільший приріст використання отримують **Vitest** і **Playwright**, які додали **14** відсоткових пунктів за рік. 


      Водночас **Rolldown** показав величезне зростання впізнаваності назви — з **33%** до **51%**. А якщо дивитися на вдоволеність, то **Next.js** зазнав найбільшого падіння — з **68%** до **55%**. 


      Також цікаво бачити, що і **Preact**, і **React** значно краще оцінені респондентами, які направду ними користувалися, ніж тими, хто лише чув про них (як показано у вкладці "Схвалення").

  - key: libraries.scatterplot_overview.takeaway.js2025
    t: >
      Завдяки додаванню цього року серверних фреймворків, **Express** приєднався до **Vite**, **React** і **Jest** у бажаному верхньому правому квадранті. Але **Vue.js**, **Storybook** і **Vitest** теж зовсім поруч!

  # front_end_frameworks

  - key: front_end_frameworks.front_end_frameworks_ratios.takeaway.js2025
    t: >
      For all the talk of the front-end ecosystem being in constant turmoil, things have been remarkably stable over the past year, with no change in usage rankings apart from **Alpine.js** and **HTMX** switching places.


      The same is true when looking at satisfaction, with the exception of **Qwik** continuing to lose ranks.


      Another place where that lack of new hype is apparent is the Interest tab, with no framework apart from **Solid** and **Preact** gaining ground since last year.

  - key: front_end_frameworks.front_end_frameworks_others.takeaway.js2025
    t: >
      Setting aside **Astro**, which doesn't quite belong to the front-end framework category, and **Ember**, which has been around for a while now, the most exciting new entrant here has to be **Ripple**, taking on the tough job of being the new kind on the front-end block.

  - key: front_end_frameworks.front_end_frameworks_cardinalities.takeaway.js2025
    t: >
      The average respondent has only used **2.6** different front-end frameworks over their career, again showing that the image of the burned-out web developer jumping from framework to framework on a monthly basis isn't quite accurate anymore.

  - key: front_end_frameworks.front_end_frameworks_happiness.takeaway.js2025
    t: >
      While in slight decline, front-end happiness has been remarkably stable over the years.

  - key: front_end_frameworks.front_end_frameworks_pain_points.takeaway.js2025
    t: >
      Being the dominant framework also means **React** tops the list of things front-end developers are struggling with. 


      Other recurring issues include poor performance, state management, and just the sheer complexity of the modern front-end ecosystem.

  # meta_frameworks

  - key: meta_frameworks.meta_frameworks_ratios.takeaway.js2025
    t: >
      **Next.js** keeps on gaining ground and dominating the meta-framework category, yet paradoxically, it's also losing satisfaction at the exact same time. This is not a good sign, as its satisfaction ratio now has a **39%** gap with leader **Astro**.


      And one look at the Awareness chart is enough to see how dangerous of a competitor Astro really is, going from virtually unknown to serious contender in the span of just a few years.

  - key: meta_frameworks.meta_frameworks_others.takeaway.js2025
    t: >
      Many respondents questioned the absence of **TanStack Start** from the options offered in the survey. The reason is simply that at survey launch, the framework wasn't out of beta yet!


      Thankfully respondents were able to rectify this omission by mentioning the project in the “other meta-frameworks” field, a clear signal that TanStack Start is going to make a very strong entrance in next year's ”official” rankings.

  - key: meta_frameworks.meta_frameworks_cardinalities.takeaway.js2025
    t: >
      Respondents haven't used that many different meta-frameworks, which is to be expected in a category still dominated by just one or two solutions.

  - key: meta_frameworks.meta_frameworks_happiness.takeaway.js2025
    t: >
      It will be interesting to see if diminishing satisfaction with Next.js eventually translates into diminishing happiness overall, but so far at least that doesn't seem to be the case.

  - key: meta_frameworks.front_end_frameworks_pain_points.takeaway.js2025
    t: >
      By their nature, meta-frameworks straddle the gap between client and server, which leads to a lot of complex issues. 


      And while **Next.js** offers quite a few ways to address those issues, the downside of all that power is that its learning curve and overall developer experience can sometimes suffer.

  # back_end_frameworks

  - key: back_end_frameworks.back_end_frameworks_ratios.takeaway.js2025
    t: >
      After three years of absence from the survey, it was high time we checked back on back-end frameworks. And no surprise here, **Express** is still the leader in terms of usage, even if **Nest** is progressing at a nice clip.


      But things are more interesting when switching to the Satisfactiont tab, where the top three projects (**Hono**, **Nitro**, and **ElysiaJS**) are all new entrants, making this once of the most exciting categories to watch in years to come.

  - key: back_end_frameworks.back_end_frameworks_others.takeaway.js2025
    t: >
      **oRPC** offers typesafe APIs, which is a big selling point compared to traditional back-end solutions created before the TypeScript sea change.

  - key: back_end_frameworks.back_end_frameworks_cardinalities.takeaway.js2025
    t: >
      Respondents tend to use relatively few back-end frameworks, which might be why **Express** remains so dominant: once you've picked a back-end, you're unlikely to go out of your way to try out new options.

  - key: back_end_frameworks.back_end_frameworks_pain_points.takeaway.js2025
    t: >
      While **Express* is still widely used *15 years* after its initial release, some are starting to wish the back-end ecosystem would move on to more modern alternatives. 


      This is especially true when it comes to providing typesafe, TypeScript-compatible APIs, which is something that venerable framework was never meant to do.

  # testing

  - key: testing.testing_ratios.takeaway.js2025
    t: >
      While **Jest** has long enjoyed a comfortable spot as the most-used testing tool, **Vitest** is climbing the ranks so fast that it wouldn't be surprising to see it overtake it in the upcoming year, especially with Jest's satisfaction ratio trending down. 


      Meanwhile, **Storybook** is an interesting case study in a project trying its best to address its users' frustrations. It's still showing a slight satisfaction loss compared to last year, but the rate of decline has slowed, and things might start looking up soon.

  - key: testing.testing_others.takeaway.js2025
    t: >
      While **Deno**'s test runner wasn't included in the survey's main options, it did come first in freeform mentions.

  - key: testing.testing_cardinalities.takeaway.js2025
    t: >
      While other categories so far have hovered around 1-2 tools used, when it comes to testing respondents averaged a whopping **4.4**, a sign that people may still be looking for that perfect test runner, and that things are far from settled in this area.

  - key: testing.testing_happiness.takeaway.js2025
    t: >
      Happiness remains stable, with a slight uptick over the past three years.

  - key: testing.testing_pain_points.takeaway.js2025
    t: >
      **Mocking** remains a major pain point for developers, with **configuration** and setup not far behind.

  # build_tools

  - key: build_tools.build_tools_ratios.takeaway.js2025
    t: >
      I really thought this would be the year **Vite** finally overtakes **webpack** to claim the spot of most-used build tools, but it looks like this might have to wait one more year. 


      **Turbopack** is also showing a lot of progress in terms of usage, no doubt benefiting from being bundled with Next.js. That being said, similar to Next.js its satisfaction ratio is quite worrying.


      **Rolldown** on the other hand has gained satisfaction, and also comes first in terms of interest.


      Finally, **Bun** is doing quite well across all categories, which is quite an achievement for such a new and ambitious project.

  - key: build_tools.build_tools_others.takeaway.js2025
    t: >
      Once again respondents mentioned **Deno**, making a strong case to include it in the main survey options in 2026.

  - key: build_tools.build_tools_cardinalities.takeaway.js2025
    t: >
      Similar to testing tools, respondents have been trying out quite a few different build tools. It remains to be seen if **Vite** will make this a thing of the past.

  - key: build_tools.build_tools_happiness.takeaway.js2025
    t: >
      Build tools happiness remains stable.

  - key: build_tools.build_tools_pain_points.takeaway.js2025
    t: >
      While Vite has certainly made things better, **configuration** and setup remain the most pressing issue when it comes to build tools.

    # other_tools

  - key: other_tools.libraries.takeaway.js2025
    t: >
      **Zod** saw a lot of growth year-over-year. And since many of the libraries featured here focus on date management, it'll be interesting to see how this chart evovles as **Temporal** becomes better-supported across browsers.

  - key: other_tools.graphics_animation.takeaway.js2025
    t: >
      **Chart.js** was a write-in result last year, and being included as an official option this survey rocketed it all the way up to first place.

  - key: other_tools.utilities.takeaway.js2025
    t: >
      Every self-respecting JavaScript developer these days seems to be using **ESLint** and **Prettier**.

  - key: other_tools.runtimes.takeaway.js2025
    t: >
      **Node.js** and browsers taking the top spots isn't a big shock, but **Bun**'s third place is a nice sign of the project's growth.

  - key: other_tools.hosting.takeaway.js2025
    t: >
      Despite the company's sometimes-polarizing reputation, **Vercel** has done a pretty amazing job of establishing itself as a serious hosting option, second only to **AWS**.

  - key: other_tools.edge_runtimes.takeaway.js2025
    t: >
      While **Cloudflare** was fifth in the hosting chart, it ranks just below **AWS Lambda** here, making it a serious contender in both spaces.

  - key: other_tools.text_editors.takeaway.js2025
    t: >
      **VS Code** has long dominated the text editor ranks, but **Cursor** has somehow managed to outpace every other competitor thanks to its promise of AI-powered coding.

  - key: other_tools.mobile_desktop.takeaway.js2025
    t: >
      **React Native** takes the lead here, but **Expo** and **Tauri** are also worth keeping an eye on!

  - key: other_tools.monorepo_tools.takeaway.js2025
    t: >
      **pnpm** is the clear leader in the monorepo category.

  - key: other_tools.non_js_languages.takeaway.js2025
    t: >
      When we're not writing JavaScript, most of us turn to **Python** and **PHP**.

  - key: other_tools.non_js_languages_features_top_features.takeaway.js2025
    t: >
      This year we also asked respondents which *features* they would like to see adopted in JavaScript, while also specifying the language that feature came from.


      The second tab shows which languages appeared most frequently. And while **Rust** was only the 7th-most used language in the previous “Non-JavaScript Languages” chart, it came in first here, a testament to its outsized influenced on the JavaScript ecosystem.

  # usage

  - key: usage.js_ts_balance.takeaway.js2025
    t: >
      The proportion of developers who write the entirety of their code in TypeScript keeps increasing, and may soon represent a majority of respondents.

  - key: usage.compiled_code_balance.takeaway.js2025
    t: >
      A natural consequence of writing typed code (at least today) is depending on a build step, which is the case for the vast majority of us.

  - key: usage.build_step_advantages.takeaway.js2025
    t: >
      If we *did* want to envision a build-less future, the first step would no doubt be native type support in JavaScript.

  - key: usage.ai_generated_code_balance.takeaway.js2025
    t: >
      While respondents do not yet produce a majority of their code with AI, the balance is quickly shifting, with code being the result of AI output increasing from 20% to 29% year-over-year.

  - key: usage.what_do_you_use_js_for.takeaway.js2025
    t: >
      Despite the ever-growing number of use cases for JavaScript, front-end develeopment remains the most common reason to use the language for now.

  - key: usage.industry_sector.takeaway.js2025
    t: >
      Given the survey's focus on upcoming trends and early-stage features, it makes sense that its largest audience block would be developers building tools and products for other programmers.

  - key: usage.js_app_patterns.takeaway.js2025
    t: >
      Modern JavaScript frameworks may now support a whole range of various rendering modes, but it turns out the simplest ones remain the most popular to this day.

  - key: usage.top_js_pain_points.takeaway.js2025
    t: >
      We now have pretty much all the tools we need; the challenge is to know how to use them right.

  - key: usage.state_of_the_web_happiness.takeaway.js2025
    t: >
      Happiness is once again stable.

  - key: usage.state_of_js_happiness.takeaway.js2025
    t: >
      Happiness levels remain constant once again.

  # resources

  - key: resources.courses.takeaway.js2025
    t: >
      It's always interesting to see how independent developers such as **Wes Bos** and **Fireship** have managed to leverage their following into competing with much larger entities such as **Udemy** or **O'Reily**.

  - key: resources.newsletters.takeaway.js2025
    t: >
      Newsletters lack the built-in sharing mechanisms of social media, so growing them can be quite tough. This explains why incumbents that have been around for a long time such as **JavaScript Weekly** and **CSS-Tricks** are still at the top of the rankings.

  - key: resources.podcasts.takeaway.js2025
    t: >
      As the line between podcasts and video content becomes ever-blurrier, **Syntax** still maintains their spot as respondents' favorite programming-related audio source.

  - key: resources.video_creators.takeaway.js2025
    t: >
      While these rankings can be very dependent on whether video creators shared the survey with their own audience or not, **Fireship**, **Theo**, and **The Primeagen** seem firmly entranched as the top three figures of programming YouTube.

  - key: resources.people.takeaway.js2025
    t: >
      While **Midudev** has his own large video presence and other “famous” developers are associated with successful open-source projects, it's worth noting that **Josh Comeau**'s well-deserved success comes down almost entirely to the quality of his excellent blog, a throwback to a more old-school way of creating instructional content online.

  - key: resources.other_surveys.takeaway.js2025
    t: >
      While filling out too many online surveys can do doubt lead to the dreaded “survey fatigue” (talk to your doctor!), when consumed in moderation they can still be a great tool to learn about the ecosystem.

  ###########################################################################
  # Picks
  ###########################################################################

  - key: picks.my_pick.js2025
    t: "Мій вибір 2025: "
  - key: picks.intro
    t: Ми покликали спільноту JavaScript поділитися своїм “вибором року”

  - key: picks.aurora_scharff.name
    t: |
      Talk: Next.js at the speed of Bun
  - key: picks.aurora_scharff.bio
    t: Web Developer
  - key: picks.aurora_scharff.description
    t: |
      My pick is Lydia Hallie's talk at Next.js Conf about Bun. They're working on super exciting things, and shes the best at explaining advanced topics :)

  - key: picks.ankita_kulkarni.name
    t: |
      Delba Oliveira
  - key: picks.ankita_kulkarni.bio
    t: Tech Educator
  - key: picks.ankita_kulkarni.description
    t: |
      My pick of the year is Delba Oliveira from the Next.js team. Her work across the ecosystem is remarkable. She has a rare ability to make complex topics feel simple while consistently delivering deeply practical guidance for modern Next.js, React and JS.

  - key: picks.shaundai_person.name
    t: |
      Bree Hall
  - key: picks.shaundai_person.bio
    t: Senior Software Engineer at Netflix
  - key: picks.shaundai_person.description
    t: |
      My pick is Bree Hall, she's been doing great work and we're lucky to have her as part of the community.

  - key: picks.sam_selikoff.name
    t: |
      RSC Explorer
  - key: picks.sam_selikoff.bio
    t: Next.js team
  - key: picks.sam_selikoff.description
    t: |
      Server Components extend the capabilities of server-rendered React by allowing you send not only inert HTML to the browser, but also data structures (objects, arrays, dates, maps), Promises (even recursive ones), new React trees (and their source code), and Server Function references. It's a serialization format, and Dan's RSC Explorer really helps you understand that.

  - key: picks.sacha_greif.name
    t: |
      Kit Langton
  - key: picks.sacha_greif.bio
    t: Devographics surveys creator
  - key: picks.sacha_greif.description
    t: |
      Kit's site is an awesome showcase of projects combining beautiful visuals, great functionality, and amazing content especially around [TypeScript](https://types.kitlangton.com/) and the [Effect](https://www.effect.institute/) library.

  - key: picks.declan_chidlow.name
    t: |
      Web Origami
  - key: picks.declan_chidlow.bio
    t: Front-end developer, designer, writer
  - key: picks.declan_chidlow.description
    t: |
      Describing itself as “a concise dialect of JavaScript expressions”, Web Origami is a project that hit its stride in 2025. It offers a remarkably intuitive way to transform data, build websites, and call JavaScript directly from the shell. It connects expressions and paths to enable powerful, intuitive data manipulation.

  - key: picks.jenn_schiffer.name
    t: |
      The Atproto Dev Community
  - key: picks.jenn_schiffer.bio
    t: Reknowned candle-maker and karaoke professional
  - key: picks.jenn_schiffer.description
    t: |
      I think the atproto dev community deserves a spotlight - one of the official SDKs is TypeScript! Also shout-out to [Jason Lengstorf](https://jason.energy/) and everyone else involved in [this](https://bsky.app/profile/codetv.dev/post/3lzgoi3nzh223).

  - key: picks.michael_rambeau.name
    t: |
      SST
  - key: picks.michael_rambeau.bio
    t: Creator, [BestOfJS](https://bestofjs.org/)
  - key: picks.michael_rambeau.description
    t: |
      I didn't know about SST just a year ago, but I started using it for a new work project and it's been fanstastic! It really deserves more exposure.

  - key: picks.adam_argyle.name
    t: |
      Observable
  - key: picks.adam_argyle.bio
    t: Co-host, The CSS Podcast
  - key: picks.adam_argyle.description
    t: |
      Observable lets you build interactive documents treating each block as a function that yields something. It's all async, super capable, fun as hell to make stuff in, and powers [Nerdy Notebook](http://nerdy.dev/notebook).

  ###########################################################################
  # Highlights
  ###########################################################################

  - key: highlight.solid.html2025
    t: |
      **Solid** may only be used by 10% of respondents, but the fact that it's had the highest satisfaction for *five years running* should be enough to make us pay attention to what it's doing. 


      What's more, Solid creator [Ryan Carniato](https://dev.to/ryansolid) has been in large part responsible for popularizing concepts such as **signals**, and is one of the best at [explaining the intricacies](https://dev.to/this-is-learning/javascript-frameworks-heading-into-2026-2hel) of other front end frameworks, too.

  - key: highlight.astro.html2025
    t: |
      **Astro** has taken a good look at what other static site generators were doing, taken the best ideas, and discarded the rest. The result is a framework that both feels simple and powerful at the same time, and has the top satisfaction ranking as a result. 


      And having recently [joined forces with Cloudflare](https://blog.cloudflare.com/astro-joins-cloudflare/), it's exciting to think about what the Astro team will show us next!

  - key: highlight.elysia.html2025
    t: |
      There are many back-end frameworks around, but **Elysia** stands out as the only one being built on top of [Bun](https://bun.com/). 


      Leveraging a brand new runtime can be a risky bet, but given Bun's own growing popularity it looks like that bet might pay off. If anything, it's certainly proving to be an interesting experiment!

  - key: highlight.bun_test.html2025
    t: |
      After years of hoping for a more unified toolchain, it looks like JavaScript developers might finally be starting to get their wish. 


      **Bun test** is just one example of a test runner being baked into the runtime, with [Node](https://nodejs.org/api/test.html) and [Deno](https://docs.deno.com/runtime/fundamentals/testing/) offering their own equivalents.

  - key: highlight.rolldown.html2025
    t: |
      While the **Vite** team is mostly known for, well, Vite, it's also been quietly developing **Rolldown** as a faster, Rust-based bundler. 


      And now that [Vite itself uses Rolldown](https://vite.dev/blog/announcing-vite8-beta), the project will no doubt take a central role in the build tool ecosystem.

  ###########################################################################
  # Awards
  ###########################################################################

  - key: award.tool_usage_delta_award.comment
    t: |
      І **Vitest**, і **Playwright** мають частку використання **{value}** серед респондентів.

  - key: award.tool_satisfaction_award.comment
    t: Vite знову посідає перше місце — **{value}** розробників ставляться до нього позитивно.

  - key: award.tool_interest_award.comment
    t: Знову цього року **Vitest** зацікавив найбільшу частку розробників, а саме **{value}**.

  - key: award.most_write_ins_award.comment
    t: "**fnm**, менеджер версій для Node, вкотре доводить, що краще те, що написано мовою Rust; кількість згадувань — **{value}**."

  - key: award.most_commented_tool_award.comment
    t: "**Next.js** точно не дає нікому спокою; кількість коментарів — **{value}**."

  # - key: award.most_loved_tool_award.comment
  #   t: With **{value}** of all developers having a positive opinion of it, Vite once more gets the prize for most-loved technology!

  # - key: award.feature_adoption_delta_award.comment
  #   t: With a **{value}** progression in 2022, **top-level await** has quickly become an integral part of JavaScript.

  ###########################################################################
  # Conclusion
  ###########################################################################

  - key: conclusion.js2025.daniel_roe
    t: |
      <span class="first-line">TypeScript has won.</span>

      Not as a bundler &ndash; but as a language. Deno and Bun (have) supported it for a long time. And now, type stripping means you can write it natively in stable Node.js versions. If types as comments makes it, you'll even be able to write in the browser.

      …and the day of Vite has come. This year, we saw Vite downloads surpass webpack's. And alongside it, Vitest usage has surged upwards. It seems like a good time to opt in to the new Vite toolchain, and 2026 promises to be the year that happens &ndash; with the stable release of Rolldown to power a new, faster Vite, as well as the promise of an all-in-one 'Vite+'.

      Our tools are better than ever. Yet the existential question we are all asking is quite different: what about AI?

      AI is poised to change how we consume documentation, write code, make architectural choices, and more. Companies are betting everything on new patterns of development. For the vast majority of us, coding agents are changing how we interface with our code.

      Is that a good thing?

      Apparently by the end of 2025, just under 30% of code was AI-generated. Cursor and Zed are surging in popularity, even if they are far from dethroning VSCode as the most-used IDE. And agent-based approaches, like Claude, Gemini and Copilot are growing quickly.

      For those of us who are developers, knowing what 'good' looks like will be more important than ever, whatever tools we use. Keeping on top of new language features will be more important. As will knowing what libraries to build on, rather than vibe coding everything from scratch.

      It's already possible to spin up a new project in a day &ndash; or migrate an existing project easily. That's a challenge for framework and library authors. We need to make sure our tools continue to serve developers. We can't count on inertia.

      And that's a reason for all users to be optimistic.

      Let's see what 2026 brings. I'm hoping it will be faster tools, better developer experience, and the power of technology acting as a force multiplier to augment our own agency and choices.

  - key: conclusion.js2025.daniel_roe.bio
    t: Open source maintainer and founder, leading the Nuxt core team.

  ###########################################################################
  # FAQ/About
  ###########################################################################

  - key: about.content.js2025
    t: |
      Опитування "Cтан JavaScript 2025" тривало з 24 вересня до 10 листопада 2025 року та зібрало 13002 відповіді. Результати були опубліковані 3 лютого 2026 року. Опитування проводив [Devographics](https://devographics.com/). 

      Логотип і футболку "Стан JavaScript" розробив [Christopher Kirk-Nielsen](http://chriskirknielsen.com/). 

      ### Цілі опитування 

      Це опитування було створено для визначення майбутніх тенденцій в екосистемі веброзробки, щоб допомогти розробникам здійснити вибір з-поміж наявних технологій. 

      Так опитування зосереджено більше на передбаченні того, що станеться протягом наступних кількох років, ніж на аналізі того, що популярно саме зараз, — ось чому деякі функції чи технології іноді можуть не згадуватися попри широке впровадження. 

      Його також слід сприймати як **знімок певної підмножини розробників у певний момент часу**, а не промовистий опис від імені всієї екосистеми. 

      ### Розробка опитування 

      Опитування було розроблено частково на основі торішнього опитування з відкритим періодом для зворотного зв'язку, під час якого план опитування обговорювався [на платформі GitHub](https://github.com/Devographics/surveys/issues/255). 

      Усі запитання опитування були необов'язковими. 

      ### Аудиторія опитування 

      Опитування було повністю відкрите та доступне онлайн, респонденти жодним чином не відфільтровувалися та не відбиралися. Респонденти складалися здебільшого з учасників минулих опитувань (сповіщених електронною поштою за спеціальним списком розсилання) та тих, хто прийшов через соціальні мережі. 

      Зауважте, що фреймворки, бібліотеки тощо могли вільно заохочувати свою аудиторію до участі в опитуванні, і деякі з них скористалися цим за допомогою банерів, твітів або в інший спосіб. Детальний розподіл джерел трафіку включено в діаграму "Джерела". 

      ### Фінансування проєкту 

      Крім як із продажу футболок, цей проєкт фінансується такими спонсорами:

      - [Google](https://www.google.com/chrome/?utm_source=stateofjs&utm_medium=website&utm_campaign=js2025)
      - [TokyoDev](https://www.tokyodev.com/?utm_source=stateofjs&utm_medium=website&utm_campaign=js2025)
      - [Frontend Masters](https://frontendmasters.com/?utm_source=stateofjs&utm_medium=website&utm_campaign=js2025)
      - [Algolia](https://www.algolia.com/?utm_source=stateofjs&utm_medium=website&utm_campaign=js2025)
      - [Storyblok](https://www.storyblok.com/?utm_source=stateofjs&utm_medium=website&utm_campaign=js2025)
      - [JetBrains](https://www.jetbrains.com/webstorm/?utm_source=stateofjs&utm_medium=website&utm_campaign=js2025)

      Будь-який внесок або спонсорство дуже цінується. Особливо ми хотіли б тісніше співпрацювати з більшою кількістю постачальників браузерів, оскільки вони відіграють центральну роль в екосистемі вебу.  

      ### Технічний огляд 

      Більш детальний технічний огляд того, як проводяться опитування, [тут](https://dev.to/sachagreif/how-the-devographics-surveys-are-run-2023-edition-1p6a).