Cursor
The AI code editor that makes plain VS Code feel outdated
Affiliate link — I may earn a commission at no extra cost to you. Visit official site →
What is Cursor?
Cursor is a fork of VS Code rebuilt around AI. It keeps everything developers already know — extensions, themes, keybindings — and layers in deep AI: Tab to autocomplete whole blocks, Cmd/Ctrl+K to edit code in place, and an Agent/Chat mode that can plan and apply changes across many files at once.
What sets it apart is codebase awareness: Cursor indexes your whole project so its suggestions actually fit your code, not generic boilerplate. For most developers it is the fastest way to ship today, and it has become the default editor for a large slice of the AI-coding crowd.
✓ Benefits
- Codebase-aware suggestionsCursor indexes your entire repo, so completions and chat answers reference your real functions, types and patterns instead of inventing generic code.
- Multi-file edits from one promptAgent mode can plan a change, touch several files, and show a reviewable diff — a huge speed-up for refactors and feature work.
- Best-in-class Tab autocompleteThe next-edit prediction often guesses your exact intent across multiple lines, so you accept whole edits with one keystroke.
- Zero learning curve for VS Code usersIt is VS Code underneath — your extensions, settings and shortcuts carry over, so you are productive on day one.
- Model choice and flexibilityYou can switch between frontier models (Claude, GPT, Gemini) for different tasks, and bring your own API key if you prefer.
✕ Disadvantages
- Pro pricing adds up$20/mo is fine for full-time devs, but freelancers and hobbyists can find the fast-request limits restrictive, pushing them toward heavier usage costs.
- Confidently wrong suggestionsLike any AI tool it sometimes proposes plausible-looking code that is subtly broken. You still need to read and test every change.
- Slows down on very large reposIndexing and agent runs can lag on huge monorepos, and context limits mean it occasionally misses relevant files.
- Easy to over-rely on itAccepting edits without understanding them can erode your own grasp of the codebase and let bugs slip through review.
Key features
- ▹Codebase-wide indexing and context
- ▹Tab next-edit autocomplete
- ▹Cmd/Ctrl+K inline editing
- ▹Agent mode for multi-file changes
- ▹Choice of frontier AI models
- ▹Full VS Code extension compatibility
- ▹Bring-your-own API key option
- ▹Built-in chat with @-file references
Best for
- ★Building features faster across multiple files
- ★Refactoring and modernising legacy code
- ★Writing tests and boilerplate in seconds
- ★Learning an unfamiliar codebase by asking it questions
- ★Debugging with full-project context
Cursor Pricing
- ✓2-week Pro trial
- ✓Limited fast requests
- ✓Tab autocomplete
- ✓Basic AI features
- ✓Generous fast requests
- ✓Unlimited slow requests
- ✓Agent multi-file edits
- ✓Frontier model access
- ✓Everything in Pro
- ✓Centralised team billing
- ✓Admin dashboard
- ✓Privacy mode enforced
Affiliate disclosure: This review contains affiliate links. If you sign up through my links, I may earn a commission at no additional cost to you. This helps support the independent reviews on tnx.ai. I only recommend tools I've personally tested and believe are worth your money. Full disclosure policy →