Cursor
A code editor built from the ground up for AI-assisted coding.
Description
Cursor is a code editor (forked from VS Code) with deep AI integration — chat with your codebase, auto-fix bugs, and refactor across multiple files at once.
Pros
- Strong understanding of your whole codebase
- Smooth experience since it's built for AI coding
- Ships new features quickly
Cons
- Takes adjustment if you're used to plain VS Code
- Free tier limited on AI requests, and the credit system takes a bit to understand
What it does & solves
Cursor solves the friction of switching between your editor and a separate AI chat window. Because it's a full editor with the model wired directly into the codebase index, it can answer questions about your whole project, apply multi-file edits through its Composer/Agent mode, and catch its own mistakes by running your tests or linter. It's aimed at developers who want AI woven into every part of the workflow rather than bolted on as a plugin.
Use cases
- Asking chat questions about how a feature works across your whole codebase
- Using Agent/Composer mode to implement a feature spanning several files
- Auto-fixing bugs flagged by the linter or a failing test
- Refactoring a function or module and having related call sites updated automatically
- Reviewing an AI-generated diff before merging
- Migrating a small codebase or config format with guided, reviewable edits
- Onboarding onto an unfamiliar repo faster by asking Cursor to explain its structure
Who it's for
Built for developers already comfortable in VS Code who want AI as a first-class part of the editor rather than an add-on extension — from solo builders and freelancers to engineering teams on the Business plan. Because it's a full VS Code fork, existing extensions and keybindings mostly carry over, so the learning curve is more about new AI-specific workflows (Composer, Agent, rules files) than the editor itself.
Getting started
- 1Download Cursor and sign in — it can import your VS Code settings, extensions, and keybindings automatically.
- 2Open a project folder and let Cursor index the codebase so chat and Agent mode have full context.
- 3Ask a question or describe a task in the chat panel, or open Composer/Agent mode for multi-file changes.
- 4Review the proposed diff, accept or edit it file by file, then run your tests before committing.
- 5Add a `.cursor/rules` file once you notice yourself repeating the same instructions in prompts.
Tips for using it well
- Use Auto mode for everyday requests and only switch to a manually-selected frontier model for genuinely hard problems — model choice is what drives your premium request usage.
- Add a `.cursor/rules` file describing your stack and conventions so every AI edit follows the same patterns without you repeating them in each prompt.
- Keep prompts scoped to a feature or bug rather than 'improve this codebase' — narrow asks produce diffs you can actually review.
- Use @-mentions to point the AI at specific files or docs instead of relying on it to guess what's relevant.
- Review and stage Agent-mode diffs like you would a colleague's PR — don't accept multi-file changes wholesale without reading them.
Plans & pricing
Free plan: The free Hobby tier includes about 2,000 completions a month and 50 slower premium requests, using a basic model with context limited to the current file — codebase-wide indexing and Composer are Pro-only.
Hobby
$10/month
A lighter paid step up from the free tier for occasional AI use.
Pro
$20/month
Unlimited completions plus around 500 fast premium requests a month — the usual starting point for daily use.
Pro+
$60/month
Roughly 10x the premium request allowance of Pro, with priority access and more context.
Ultra
$200/month
The highest request ceiling, built for heavy, continuous agentic coding.