Cursor
The AI code editor built for agentic, multi-file engineering work.
Cursor is a fork of VS Code with model access wired into the editing loop rather than bolted on as a sidebar. The differentiator is repo-wide context: it indexes your codebase so a request like "move auth into a middleware and update every caller" produces a reviewable multi-file diff instead of a snippet you paste yourself.
For Indian students and early-career engineers, the practical value is speed of comprehension on unfamiliar repos, ask it to explain a module, then have it write the first test. Treat generated diffs as a junior's pull request: read every line before accepting.
What it's genuinely good at
- Multi-file edits land as reviewable diffs
- Repo-wide indexing gives genuinely relevant context
- Keeps the entire VS Code extension ecosystem
- Tab completion is the best-in-class part of the product
Where it falls apart
- Indexing is heavy on large monorepos
- Best models are behind the paid tier
- Agent mode can over-edit if the prompt is vague
Alternatives to Cursor
Who should use Cursor
- Engineers shipping features in an existing codebase daily
- Anyone onboarding onto an unfamiliar repo
- Students who want to read more code than they write
What it costs
Free hobby tier with limited fast requests; Pro around $20/month unlocks frontier models and higher limits. Students should check the education discount before paying.
Frequently asked
Is Cursor worth paying for over GitHub Copilot?
If you mostly want inline completion, Copilot is cheaper. If you want multi-file refactors and codebase Q&A, Cursor's agent workflow is the stronger product.
Does Cursor send my code to a server?
Yes by default. Privacy Mode stops code from being retained, and is the setting to enable before pointing it at employer code.