Hayz Labs · https://hayzlabs.com/reviews/top/languages/
reviews · top 20 list · checked 19 Oct
Top 20 programming languages
The languages developers use most, with what each is best for and what you give up. Useful when you choose a language for a project you will build with AI help.
Nothing on this list is scored yet: no outside source we can turn into a number covers them. The order below follows the public measure named on each row.
By share of developers using it in the Stack Overflow 2026 survey (all respondents). Markup and styling languages such as HTML and CSS are left out. Contains information from the Stack Overflow Developer Survey 2026, which is made available under the Open Database License (ODbL). Open the Table to see the number that placed each one and how it was scored.
- 1JavaScriptEcma TC39
The language of web pages, also run on servers through Node.js
Best for Interactive web pages, plus servers and tools through Node.jsYou give up Only 42.0% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 2SQLISO/IEC JTC 1/SC 32 (international standard)
The standard language for asking questions of, and changing, data stored in database tables
Best for Reading and changing data in tables (rows and columns) inside a databaseYou give up Rows come back in no set order unless you sort them; 51.4% of users want to keep it
Not scored yetFree - 3PythonPython Software Foundation
General-purpose language used for web, data, science, scripting and teaching
Best for Web, data and science work, scripting, teaching, and software and game toolsYou give up Only about half of current users (50.6%) want to keep using it (Stack Overflow 2026)
Not scored yetFree - 4Bash/ShellGNU Project
The command line of Linux and Mac; scripts chain commands together to automate tasks
Best for Automating the command line: running commands and small scripts on servers and PCsYou give up Only 48.1% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 5TypeScriptMicrosoft
JavaScript with added labels for data types, so editors can catch mistakes early
Best for Larger JavaScript projects in browsers, Node.js, Deno and BunYou give up Usually converted to plain JavaScript before it runs, an extra build step
Not scored yetFree - 6JavaOracle and the OpenJDK community
Long-standing language for business back ends, run by the Java virtual machine
Best for Large business (enterprise) applications and back-end servicesYou give up Needs the JDK toolkit and runs in a virtual machine; 38.5% of users want to keep it
Not scored yetFree - 7C#Microsoft
Microsoft's main language for the .NET platform: web, desktop, phone and cloud apps
Best for Apps on .NET: web back ends (ASP.NET Core), desktop, phone and cloudYou give up In one 2026 AI coding test, a model scored 57% in C#, against about 70% in most others
Not scored yetFree - 8PowerShellMicrosoft
Microsoft's command line and scripting language for managing computers, on Windows, Linux and Mac
Best for Automating system admin tasks and build, test and deploy stepsYou give up Only 31.5% of users want to keep using it, the lowest of these 20 (Stack Overflow 2026)
Not scored yetFree - 9C++ISO C++ committee (WG21)
Fast, long-lived language for systems, embedded and other speed-critical software
Best for Systems code, embedded devices, and apps where speed per watt mattersYou give up The C++ FAQ says it takes 6 to 12 months to become broadly proficient
Not scored yetFree - 10CISO C committee (WG14)
Small, old language close to the hardware; much other software is built on it
Best for Systems programming: low-level code close to the hardwareYou give up A bad pointer (a memory address) can crash the program; 45.9% of users want to keep it
Not scored yetFree - 11RustRust Foundation
Fast systems language whose rules prevent many memory bugs before the program runs
Best for Command-line tools, network services, WebAssembly and small embedded devicesYou give up In one 2026 AI coding test, a model scored 48% in Rust, against 74% in Python
Not scored yetFree - 12PHPThe PHP Group
Server-side language mixed into web pages to build websites
Best for Web development: code on the server builds the HTML pages visitors seeYou give up Only 35.7% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 13GoGoogle
Simple, fast language from Google for servers and cloud tools
Best for Cloud and network services, command-line tools, web apps and DevOps toolsYou give up Only 49.5% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 14KotlinJetBrains
Concise language from JetBrains for Android, cross-platform and server apps; works with Java
Best for Android and cross-platform apps, plus server back endsYou give up Only 43.1% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 15LuaLua team (PUC-Rio)
Small scripting language built into other programs, especially games
Best for Scripting inside games and other apps written in C, C++ and moreYou give up Built to be embedded in another program; 40.8% want to keep using it (Stack Overflow 2026)
Not scored yetFree - 16AssemblyNo single steward (each chip family has its own)
Human-readable form of a processor's own instructions; the lowest level most people code at
Best for Very low-level code: boot code, drivers and hand-tuned speed-critical partsYou give up Only 48.3% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 17SwiftApple
Apple's language for iPhone, Mac and other Apple apps; also runs on Linux and Windows
Best for Apps for Apple devices, plus cloud services and small embedded devicesYou give up Only 48.5% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 18RubyRuby core team
Flexible scripting language best known for the Ruby on Rails web framework
Best for Web apps with Ruby on Rails, and scriptingYou give up In one 2026 AI coding test, a model scored 49% in Ruby, against 74% in Python
Not scored yetFree - 19DartGoogle
Google's language behind Flutter, for building one app for phone, desktop and web
Best for Flutter apps for mobile, desktop and web, plus some back endsYou give up Only 42.3% of current users want to keep using it (Stack Overflow 2026)
Not scored yetFree - 20RR Foundation
Language for statistics, data analysis and charts
Best for Statistics: modelling, tests, time series, clustering, and chartsYou give up Heavy number crunching is handed to C, C++ or Fortran code; 40.7% of users want to keep it
Not scored yetFree
| # | Name | Score | What placed it | Price | Best for | You give up |
|---|---|---|---|---|---|---|
| 1 | JavaScriptEcma TC39 · Open source | Not scored | Used by about 62% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'JavaScript' 62.0% (8,742 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Interactive web pages, plus servers and tools through Node.js | Only 42.0% of current users want to keep using it (Stack Overflow 2026) |
| 2 | SQLISO/IEC JTC 1/SC 32 (international standard) · Open source | Not scored | Used by about 58% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'SQL' 58.4% (8,237 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Reading and changing data in tables (rows and columns) inside a database | Rows come back in no set order unless you sort them; 51.4% of users want to keep it |
| 3 | PythonPython Software Foundation · Open source | Not scored | Used by about 58% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Python' 58.0% (8,184 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Web, data and science work, scripting, teaching, and software and game tools | Only about half of current users (50.6%) want to keep using it (Stack Overflow 2026) |
| 4 | Bash/ShellGNU Project · Open source | Not scored | Used by about 51% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Bash/Shell (all shells)' 51.3% (7,231 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Automating the command line: running commands and small scripts on servers and PCs | Only 48.1% of current users want to keep using it (Stack Overflow 2026) |
| 5 | TypeScriptMicrosoft · Open source | Not scored | Used by about 44% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'TypeScript' 43.8% (6,171 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Larger JavaScript projects in browsers, Node.js, Deno and Bun | Usually converted to plain JavaScript before it runs, an extra build step |
| 6 | JavaOracle and the OpenJDK community · Open source | Not scored | Used by about 28% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Java' 27.6% (3,885 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Large business (enterprise) applications and back-end services | Needs the JDK toolkit and runs in a virtual machine; 38.5% of users want to keep it |
| 7 | C#Microsoft · Open source | Not scored | Used by about 27% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'C#' 27.1% (3,826 of 14,099). not scored: no 2026 test of AI coding covers 20 languages; the tradeoff figure is from the Multi-LCB paper, Table 1 (https://arxiv.org/html/2606.20517v1) Report a mistake | Free | Apps on .NET: web back ends (ASP.NET Core), desktop, phone and cloud | In one 2026 AI coding test, a model scored 57% in C#, against about 70% in most others |
| 8 | PowerShellMicrosoft · Open source | Not scored | Used by about 25% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'PowerShell' 24.6% (3,472 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Automating system admin tasks and build, test and deploy steps | Only 31.5% of users want to keep using it, the lowest of these 20 (Stack Overflow 2026) |
| 9 | C++ISO C++ committee (WG21) · Open source | Not scored | Used by about 24% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'C++' 24.3% (3,425 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Systems code, embedded devices, and apps where speed per watt matters | The C++ FAQ says it takes 6 to 12 months to become broadly proficient |
| 10 | CISO C committee (WG14) · Open source | Not scored | Used by about 22% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'C' 21.8% (3,070 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Systems programming: low-level code close to the hardware | A bad pointer (a memory address) can crash the program; 45.9% of users want to keep it |
| 11 | RustRust Foundation · Open source | Not scored | Used by about 18% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Rust' 18.0% (2,542 of 14,099). not scored: no 2026 test of AI coding covers 20 languages; the tradeoff figure is from the Multi-LCB paper, Table 1 (https://arxiv.org/html/2606.20517v1) Report a mistake | Free | Command-line tools, network services, WebAssembly and small embedded devices | In one 2026 AI coding test, a model scored 48% in Rust, against 74% in Python |
| 12 | PHPThe PHP Group · Open source | Not scored | Used by about 18% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'PHP' 17.7% (2,490 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Web development: code on the server builds the HTML pages visitors see | Only 35.7% of current users want to keep using it (Stack Overflow 2026) |
| 13 | GoGoogle · Open source | Not scored | Used by about 17% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Go' 16.5% (2,332 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Cloud and network services, command-line tools, web apps and DevOps tools | Only 49.5% of current users want to keep using it (Stack Overflow 2026) |
| 14 | KotlinJetBrains · Open source | Not scored | Used by about 11% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Kotlin' 10.8% (1,524 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Android and cross-platform apps, plus server back ends | Only 43.1% of current users want to keep using it (Stack Overflow 2026) |
| 15 | LuaLua team (PUC-Rio) · Open source | Not scored | Used by about 9% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Lua' 9.4% (1,327 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Scripting inside games and other apps written in C, C++ and more | Built to be embedded in another program; 40.8% want to keep using it (Stack Overflow 2026) |
| 16 | AssemblyNo single steward (each chip family has its own) · Open source | Not scored | Used by about 7% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Assembly' 7.2% (1,020 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Very low-level code: boot code, drivers and hand-tuned speed-critical parts | Only 48.3% of current users want to keep using it (Stack Overflow 2026) |
| 17 | SwiftApple · Open source | Not scored | Used by about 6% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Swift' 5.9% (831 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Apps for Apple devices, plus cloud services and small embedded devices | Only 48.5% of current users want to keep using it (Stack Overflow 2026) |
| 18 | RubyRuby core team · Open source | Not scored | Used by about 6% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Ruby' 5.8% (823 of 14,099). not scored: no 2026 test of AI coding covers 20 languages; the tradeoff figure is from the Multi-LCB paper, Table 1 (https://arxiv.org/html/2606.20517v1) Report a mistake | Free | Web apps with Ruby on Rails, and scripting | In one 2026 AI coding test, a model scored 49% in Ruby, against 74% in Python |
| 19 | DartGoogle · Open source | Not scored | Used by about 5% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'Dart' 4.9% (697 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Flutter apps for mobile, desktop and web, plus some back ends | Only 42.3% of current users want to keep using it (Stack Overflow 2026) |
| 20 | RR Foundation · Open source | Not scored | Used by about 5% of developers asked (Stack Overflow survey 2026)NotesTheir figure: SO 2026 'R' 4.5% (639 of 14,099). not scored: no 2026 test of AI coding covers 20 languages Report a mistake | Free | Statistics: modelling, tests, time series, clustering, and charts | Heavy number crunching is handed to C, C++ or Fortran code; 40.7% of users want to keep it |
Nothing matches all of that. , or try everything.
How we ranked this
Order. By share of developers using it in the Stack Overflow 2026 survey (all respondents). Markup and styling languages such as HTML and CSS are left out. Contains information from the Stack Overflow Developer Survey 2026, which is made available under the Open Database License (ODbL).
How we score this list. Not scored: no independent test from 2026 measures how well AI models write all 20 languages. We will score them when one does. The full rules are on How we score. Open "How it was scored" on any row to see the working.
What we did not do. We did not run any of these ourselves. Every number comes from the public source linked on its row, checked on 19 Oct 2026.
Spot a mistake? A wrong rank, price or link: tell us and we will check it against the source. Corrections are dated in News.
Reviews are drafted with help from AI models, including Anthropic's Claude. Model scores come only from outside leaderboards and independent evaluations, by a written rule.