
Compiled vs Interpreted Languages Explained
Short answer: a compiled language is translated into machine code before the program runs, which usually gives faster execution. An interpreted language is read and executed line by line by another program (the interpreter), which makes development and portability easier. Many modern languages mix both ideas.
How compilation works
A compiler reads all of your source code and produces an executable file for a specific kind of computer. You run the executable without the compiler.
gcc hello.c -o hello
./hello
Examples: C, C++, Go and Rust. The compiler also finds many errors, such as type mistakes, before the program ever runs.
How interpretation works
An interpreter reads the source code and executes it directly, step by step, each time you run the program.
python hello.py
Examples: Python, Ruby and PHP. You can change a file and run it again immediately, which makes experimenting fast.
Comparison
| Compiled | Interpreted | |
|---|---|---|
| Execution speed | Usually faster | Usually slower |
| Start-up for development | Compile step needed | Run immediately |
| Error detection | Many errors found before running | Some errors appear only when that line runs |
| Portability | Build for each platform | Runs wherever the interpreter exists |
| Typical use | Systems, games, performance-critical code | Scripting, automation, web and data work |
The reality is mixed
- Java and C# compile to bytecode, and a virtual machine runs it, using just-in-time (JIT) compilation to speed it up.
- JavaScript engines such as V8 compile hot code to machine code while the page runs.
- Python compiles source to bytecode internally and then interprets it.
So "compiled" or "interpreted" describes an implementation more than the language itself. The same language can have both kinds of implementation.
Which should you choose?
Choose based on the problem. For performance-critical systems, compiled languages are a good fit. For quick development, scripts and data work, interpreted languages are productive. For most business applications, either works, and the ecosystem, team skills and libraries matter more than the execution model.
What happens when you run a small program
With a compiled language such as C, you run a compiler first and get a separate executable. If you change one line, you compile again. With Python, you run python app.py and the interpreter starts reading the file immediately. If a line has an error, the interpreter may run the earlier lines first and fail only when it reaches the problem.
Key takeaways
- Compilers translate the whole program ahead of time, and interpreters execute it as they read it.
- Compiled code is often faster, and interpreted code is quick to change and test.
- Modern languages blend both approaches with bytecode, virtual machines and JIT compilers.
Just-in-time compilation explained
Many popular languages sit between the two extremes. Java and C# compile source code ahead of time into an intermediate bytecode. When the program runs, a virtual machine (the JVM or .NET runtime) first interprets or quickly compiles that bytecode, watches which methods run most often, and compiles those "hot" parts into optimised machine code while the program is running. JavaScript engines such as V8 do something similar directly from source. This just-in-time (JIT) approach combines quick startup with performance that can come close to ahead-of-time compiled code for long-running programs.
Popular languages and how they run
| Language | Typical execution model |
|---|---|
| C, C++, Rust, Go | Compiled ahead of time to native machine code |
| Java, Kotlin, C# | Compiled to bytecode, then JIT-compiled by a virtual machine (ahead-of-time options also exist) |
| JavaScript | Parsed and JIT-compiled by the browser or Node.js engine |
| Python (CPython) | Compiled to bytecode, then interpreted by a virtual machine |
| PHP, Ruby | Interpreted from bytecode, with optional JIT in recent versions |
| TypeScript | Transpiled to JavaScript, then runs like JavaScript |
Common misconceptions
- "A language is compiled or interpreted." Strictly, implementations are. Python has an interpreter (CPython) and a JIT implementation (PyPy); C can be interpreted by special tools.
- "Interpreted means slow for everything." Programs that spend most time waiting on networks or databases see little difference, and libraries written in C do heavy lifting for Python.
- "Compiled code is always secure." Compilation does not prevent bugs like buffer overflows; memory safety depends on the language.
- "You must choose based on speed." Developer productivity, ecosystem and team skills usually matter more.
Explore related ideas in static vs dynamic typing, what a programming language is and what the C language is.
Frequently asked questions
Is Python compiled or interpreted?
Both. The standard implementation compiles source code into bytecode and then interprets that bytecode, though it is usually called an interpreted language.
Are compiled languages always faster?
Often, but not always. Algorithms, data structures and the quality of the code usually matter more than the execution model, and JIT compilers can narrow the gap.
What is the difference between a compiler and an interpreter?
A compiler translates the whole program ahead of time into another form, while an interpreter executes the program directly as it reads it.
What is transpiling?
Transpiling converts source code from one high-level language to another, such as TypeScript to JavaScript or modern JavaScript to older JavaScript. The output is still source code, not machine code.
Is JavaScript compiled?
Modern engines compile JavaScript to machine code at runtime using JIT techniques, so it is not purely interpreted, even though you never run a separate compile step yourself.
Does the choice affect how I distribute my program?
Yes. A natively compiled program is usually shipped as a binary for each operating system and processor type, while an interpreted program needs the interpreter installed on the user's machine or bundled with the app. Languages with virtual machines, such as Java, need a compatible runtime.
