Which laptop for coding? What developers actually need in 2026

Published 27 October 2026 · prices below are checked every week

You don’t need a £3,000 workstation to write code. Python, JavaScript, Go or Rust run happily on a good 16 GB laptop. What changes the answer is what runs beside your editor: Docker, databases, phone emulators or AI models. Here is what each kind of developer really needs, and where people waste money.

Colourful JavaScript code in a text editor on a MacBook screen — Photo: Marc Mueller, via Wikimedia Commons (CC0).

16 GB minimum
32 GB for Docker and VMs
512 GB SSD or more
A Mac for iPhone apps

The 30-second answer

Web, front end, learning to code

16 GB is plenty for an editor, a local server and a browser full of documentation. A MacBook Air is silent and lasts all day; a refurbished business laptop is the budget way in.

Get: MacBook Air 13-inch M5 or a refurbished Dell Latitude 5410

Full stack, Docker, databases

Containers, a database, an IDE and Slack all at once: aim for 32 GB. On Windows that means a 32 GB laptop; on a Mac, 24 GB or more.

Get: ASUS ExpertBook P5, 32 GB

iPhone and Mac apps

Apple’s Xcode runs only on a Mac, so this one is decided for you. A MacBook Pro’s fan keeps long builds fast; the Air slows down a little when it gets hot.

Get: MacBook Pro 14-inch M5

Local AI and heavy builds

Running AI models on your own machine needs lots of fast memory: a Mac mini with 32 GB at your desk, or an NVIDIA graphics chip on Windows.

Get: Mac mini M6, 32 GB

The numbers that matter

16 GB

the minimum for coding in 2026. A plain laptop with 16 GB feels faster day to day than a nicer one with 8 GB

Memory

32 GB

the comfortable baseline once you run containers, virtual machines or several local services

Memory

512 GB

storage at least: node_modules, Docker images, phone SDKs and Python environments fill 256 GB fast

SSD

71%

of developers use Docker, in the biggest yearly survey of developers. It jumped 17 points in a year

Survey, 2025

How much memory? It depends on what runs beside your code

What you build	Memory	Why
Websites, front end, learning	16 GB	Editor, local server and browser fit easily
Python, data, notebooks	16–24 GB	Big datasets live in memory while you work
Full stack with Docker and a database	24–32 GB	Each container and database keeps its own share
Android or iPhone apps with emulators	24–32 GB	A phone emulator is a whole extra computer
Virtual machines	32 GB+	Every VM reserves memory for itself
Local AI models, game engines	32 GB+ and strong graphics	The model or the scene must fit in memory

Don’t overspend either: one former software developer who now reviews laptops says memory is what buyers waste most money on. If you don’t know you need more, you probably don’t. Last year’s models also give you most of this year’s speed for less.

Mac, Windows or Linux? The truth without the fan wars

In the biggest yearly developer survey, 49.5% work on Windows, 32.9% on macOS and 27.7% on Ubuntu Linux (many use more than one). Any of them works for web and back-end code; your tools decide the rest.

Mac

macOS is a certified Unix, so its terminal behaves like the servers your code runs on. Apple chips stay fast and quiet on battery, even while compiling. It’s the only way to build iPhone apps. The catch: memory can’t be upgraded later, so choose it on day one.

Windows

Needed for Visual Studio, Windows desktop apps in .NET, Power BI Desktop and much game development. WSL2 runs a real Ubuntu inside Windows, and VS Code opens straight into it. One developer who uses it daily says it fixed most of Windows’ headaches, with a few quirks and extra memory use.

Linux

Ubuntu’s long-term releases “just work” for most developers. Business laptops such as ThinkPads, Latitudes and EliteBooks are among the easiest to run it on, refurbished ones included.

Careful with Windows on Arm

Snapdragon laptops last long on battery, but one review team found Android Studio doesn’t run natively and its phone emulator doesn’t work at all. Check your tools first. Our Windows picks use Intel or AMD.

Local AI: what fits on your machine

Most AI coding help runs in the cloud and needs no special laptop. To run models on your own computer, the model has to fit in memory, with room left for your apps:

Memory a local AI model takes up download size in GB, before your other apps

8B model5.2 GB

14B model9.3 GB

20B model14 GB

30B model19 GB

What that means: with 16 GB, small models up to about 8B run well. 24 GB handles a 14B model comfortably, and 32 GB a 30B one. On a Mac the graphics share the main memory, which is why a 32 GB Mac mini is a popular AI box; on Windows an NVIDIA graphics chip makes models much faster, but its own memory (6 to 8 GB on the picks below) limits how big they can be.

Docker eats memory: keep it in check

Know where it goesOn a Mac or Windows, Docker Desktop runs your containers inside a virtual machine that holds on to memory, even when the containers are idle.
Set a ceilingDocker Desktop › Settings › Resources: give it what your projects need, not the default.
On Windows, cap WSL2 tooAdd a memory limit in a .wslconfig file in your user folder, then run wsl --shutdown.
Clean up every few weeksdocker stats shows what each container uses; stop what you don’t need, and docker system prune clears old images and build cache.

Our picks for developers

The all-rounder Mac

The MacBook Air M5: 16 GB, 1.23 kg, no fan, all-day battery. Apple will build it with 24 or 32 GB if you order it that way. Want more room for code? The 15-inch.

Get: MacBook Air 13-inch M5 or the 15-inch

Windows with 32 GB

The ASUS ExpertBook P5: 32 GB, 1 TB and a sharp 2560 × 1600 anti-glare screen in 1.27 kg. Its memory is built in, so 32 GB is for life.

Get: ASUS ExpertBook P5

32 GB for less

A ThinkPad T14s Gen 4 with 32 GB (built in), a famous keyboard and 1.26 kg; or the ThinkBook 14 G7, whose two memory slots take up to 64 GB later.

Get: ThinkPad T14s Gen 4, 32 GB or ThinkBook 14 G7

Desk power and local AI

A Mac mini M6 with 24 GB plus a 4K monitor costs less than a MacBook Pro, and it can stay on for builds and AI agents.

Get: Mac mini M6, 24 GB plus a 27-inch 4K monitor

NVIDIA for CUDA, AI and games

The MSI Prestige 14 AI Studio adds an entry-level RTX 3050 to 32 GB in a 14-inch laptop. At a desk, the Medion Recon P45 has an RTX 5060 and 32 GB (wired network only).

Get: MSI Prestige 14 AI Studio or the Medion desktop

Learning on a budget

A refurbished Dell Latitude 5410: 16 GB in two memory slots, so it can grow to 32 GB, and business laptops like it are among the easiest to run Linux on. Older processor, but fine for learning.

Get: Dell Latitude 5410, refurbished

4 mistakes to skip

Buying 8 GBA browser full of documentation plus an editor is already too much. A plain laptop with 16 GB beats a nicer one with 8 GB, Mac or not.
Paying for power you won’t useA 64 GB workstation for websites is wasted money. Match memory to the table above, and consider last year’s model on sale.
Ignoring the keyboard and screenYou type all day. ThinkPad keys travel about 1.5 mm, more than most thin laptops. A sharp screen shows more lines of code at once than a fuzzy 15.6-inch Full HD one.
Coding only on a small laptop screenFor long days, add a monitor at home: more code, less neck strain. A 14-inch and a monitor, or a 16-inch? → Which monitor for a Mac? →

A 27-inch monitor on a wooden desk with a small silver Mac mini beside it — A Mac mini and a 27-inch screen: a powerful desk setup for less than a MacBook Pro. Photo: Seasider53, CC BY 4.0, cropped, with the shop’s caption on the screen softened.

Our verdict

For most developers, a 16 GB MacBook Air or a good Windows laptop is enough for web, Python and back-end code. Spend on memory only when you run Docker, emulators, virtual machines or AI models side by side: then 32 GB pays for itself.

Let your tools decide the system: a Mac for iPhone apps, Windows for .NET desktop work and games, and either one, or Linux, for the web.

This week’s prices

Laptops with 32 GB

MacBooks

Desktops with 24 GB or more

Not sure which suits you? Answer four quick questions → Related: Mac mini M6: which one? → · Refurbished or new? →

Quick questions

Do I need a dedicated graphics card for coding?

No, unless you train AI models locally, work in game engines such as Unity or Unreal, or use CUDA. Built-in graphics handle everything else.

Is Linux easy to run on these laptops?

Yes, on the Intel and AMD ones. Business laptops such as ThinkPads, Latitudes and EliteBooks are among the best supported; try Ubuntu from a USB stick first to check Wi-Fi and sleep work.

Can I build iPhone apps on Windows?

Not properly. Apple’s Xcode, which builds and publishes iPhone apps, runs only on a Mac. Any current Mac runs it; for real app work, choose 16 GB or more.

Can I code on an iPad?

For learning Swift, yes, with Apple’s Swift Playgrounds. For anything else, it works as a window onto a remote computer, not as your main machine.

Is 16 GB enough for Python and data science?

For learning and most projects, yes. If your datasets run to several gigabytes, or you use notebooks and Docker together, 24 to 32 GB is more comfortable.

Try the free calculator: Estimate RAM and storage for your software and projects →