Cursor installs as a full standalone desktop editor, not a plugin you bolt onto something else, and the whole setup takes about five minutes. Because it is built on the same foundation as VS Code, your existing extensions, themes and keybindings can import on first launch, so the editor feels familiar before you have written a line. The only real decisions are which account to sign in with and which AI model you want driving the assistant.

Quick Answer

Download the Cursor installer from cursor.com for Windows, macOS or Linux, run it, accept the prompt to import your VS Code settings, then sign in to your Cursor account. Pick a model in the chat panel (Claude, GPT or Gemini class models are all selectable) and you are ready to code. The free Hobby tier needs no card to get started.

Step One: Download And Install

Grab the installer from the official site. Windows gets a standard .exe, macOS a signed and notarised .dmg, and Linux users can choose an AppImage, .deb or .rpm depending on the distribution. Run it like any normal application. Cursor keeps its own configuration folder separate from VS Code, so the two editors sit side by side without clashing if you want to keep both.

On a modern laptop the install is light, but the AI features lean on a decent processor and plenty of RAM once you start indexing large projects. If you are speccing a machine specifically for AI-heavy development workflows, the AI PC range at Evetech covers high-core, well-specced systems suited to local and cloud AI workloads.

Step Two: Import Your VS Code Setup

On first launch Cursor offers to bring across your VS Code extensions, settings, themes and keyboard shortcuts. Accepting this is the fastest path to a working environment, because everything you already rely on, from your formatter to your colour theme, carries over. If you prefer a clean slate you can skip it and configure from scratch, but most developers take the import and tweak from there.

Step Three: Sign In And Choose A Model

Create or sign in to a Cursor account when prompted. Once authenticated, your models, rules and any saved configuration become available immediately. Open the chat or composer panel and use the model picker to choose your engine. The selection is broad and rotates as new models ship: the Claude family, OpenAI's GPT class, Google's Gemini, plus Cursor's own in-house models are typically all on the menu.

Two practical notes. First, the free Hobby plan gives you a limited number of agent requests and tab completions, which is enough to judge whether Cursor fits how you work before you commit to a paid tier. Second, if you have your own provider API key, you can add it under settings, paste the key and verify it, which routes requests through your own account instead of Cursor's pooled credits.

Step Four: Open A Project And Test The Workflow

Open a folder and try the core features so you know they work. Inline edit lets you select code and describe a change in plain language. The chat panel answers questions about your codebase with the whole project as context. Tab completion suggests the next chunk as you type. Agent mode can plan and apply multi-file changes when you describe a larger task. Run each once on a small file to confirm your model selection is responding before you rely on it for real work.

A quick privacy check is worth doing here too. Cursor offers a privacy mode that prevents your code from being retained by the model providers, which matters if you work on commercial or client projects. Enable it in settings if your codebase is sensitive.

Picking The Right Machine

Cursor runs fine on a mid-range laptop, but the AI features feel snappier with more RAM and a faster processor, especially on large repositories. A modern multi-core chip and 16GB of memory is a comfortable baseline, with 32GB preferable if you keep many projects and containers open. When comparing options, the popular developer machines at Evetech give a good shortlist of well-balanced builds that handle a heavy editor and several browser tabs without stalling.

Frequently Asked Questions

Is Cursor free to use?

Yes, the Hobby tier is free and needs no card. It includes a capped number of agent requests and completions. Paid tiers (Pro and above) lift those limits for heavier daily use.

Do my VS Code extensions work in Cursor?

Most do, because Cursor is built on the same open foundation as VS Code. On first launch it offers to import your extensions, settings and keybindings, so your setup transfers with one click.

Can I use my own AI model API key?

Yes. In settings you can enable a provider, paste your own API key and verify it. Requests then run through your account rather than Cursor's included credits, which suits teams with existing provider contracts.

Which model should I choose in Cursor?

There is no single right answer. The Claude, GPT and Gemini class models all perform well for general coding. Try the same task on two of them and keep whichever gives cleaner results for your stack.

Does Cursor send my code to the cloud?

By default some context is sent to the model provider to generate responses. Enabling privacy mode prevents that data from being retained, which is the setting to use for confidential or client code.

Spec a machine that keeps Cursor and a busy browser fluid by browsing the AI PC range at Evetech, then install Cursor and start shipping faster.