I love @Cursor. It's enabled me to build (vibe code) so many web apps, sites, extensions, and little things quickly that 1. bring me joy and 2. help me with work or realize personal projects. However... I'm seeing a TON of movement around @Claude by Anthropic's Claude Code. I haven't personally tried it but it's apparently insane (and can also be expensive?) I'm curious. Should I switch? What are you currently using? Or do they both have their own use case. I right now like cursor because I can build directly in a GitHub repo or locally and it helps me learn my way around an IDE. Looking forward to hearing everyone's thoughts!
Half the threads here are about accuracy and fewer iterations. One source of rework I rarely see named: when I paste a screenshot, Cursor has to guess which element on a busy screen I actually meant, and it edits the wrong one. Then I'm re-prompting, which is its own iteration tax.
Curious how much of your rework is this specific thing versus raw model quality. And what's your fix, crop hard, or describe the element in text?
Just saw reports that SpaceX acquired Cursor in an all-stock deal valued at $60 billion.
This is one of the most insane startup outcomes we've ever seen:
Cursor was founded only ~4 years ago.
The reported price tag would make it one of the largest startup acquisitions in history.
It went from "just a VS Code wrapper" jokes to becoming the default AI coding tool for a huge number of developers.
SpaceX/xAI would instantly gain access to one of the most valuable datasets in AI: how millions of developers actually write, edit, debug, and ship code.
Reviewers largely see Cursor as a strong daily AI coding editor: it fits naturally into a VS Code-style workflow, cuts context switching, and shines at autocomplete, debugging, and multi-file edits across real codebases. Users repeatedly say it speeds up boilerplate, refactors, and feature work, while some non-coders find it approachable too. The main complaints are familiar: high memory use on large projects, occasional context loss, loops or overconfident output, frequent updates, and pricing that can feel unclear or expensive.
Cursor has significantly improved my development workflow. It feels like working with an experienced AI pair programmer that understands the project context, helps with multi-file changes, explains code, generates implementations, and assists with debugging. The seamless VS Code experience and strong code editing capabilities make it one of the best AI coding tools available today.
What needs improvement
frequent updates (2)
Occasionally the AI loses context during very large refactoring tasks, and long-running agent sessions can require manual guidance. Better memory across sessions, improved handling of massive codebases, lower resource usage, and more transparent reasoning during complex edits would make the experience even better.
I also evaluated Claude Code, GitHub Copilot, Windsurf, Codex, and VS Code with AI extensions. I chose Cursor because it offers an excellent balance of project awareness, intelligent code editing, multi-file refactoring, debugging assistance, and an interface that fits naturally into my existing development workflow.
I use Cursor daily as my primary code editor for full-stack web and backend development. It serves as an active AI pair programmer across my entire workflow—from scaffolding new modules and writing boilerplate to multi-file refactoring, debugging complex error traces, and generating targeted unit tests.
What needs improvement
Resource & Memory Usage: Cursor can become resource-intensive and consume high memory during full-codebase indexing or heavy Composer sessions on large monorepos.
Occasional "Thinking Loops" & Hallucinations: During complex multi-step edits, the agent can occasionally get stuck in loops or revert previously working logic if not guided with tight prompts.
Rapid Update Cycles: While rapid iteration is appreciated, frequent updates occasionally introduce minor UI bugs or breaking behavior with certain third-party extensions.
Pricing & Fast Requests: The pricing tier jump can be steep for individual hobbyists once fast-request allowances are exhausted.
vs Alternatives
Seamless VS Code Ecosystem: Because Cursor is built on a VS Code fork, migration is instantaneous. All extensions, keybindings, settings, and themes port over seamlessly without disrupting existing muscle memory.
Deep Context Awareness (@ symbols & codebase indexing): Cursor’s ability to index the full repository and reference specific files, folders, docs, and git commits makes its code suggestions and multi-file edits significantly more accurate than standard autocomplete tools.
Intelligent Multi-File Code Generation: Cursor Composer allows prompt-driven changes across multiple files simultaneously, handling large refactors and feature scaffolding in seconds.
Smart Autocomplete & Copilot++: The inline prediction is fast and contextual—it doesn't just guess the next line, it anticipates multi-line logic and repetitive patterns across the project.
Integrated Real-Time Debugging: Being able to highlight a terminal stack trace and ask the AI to pinpoint the exact failure line saves hours of troubleshooting.
Alternatives considered: GitHub Copilot in standard VS Code, JetBrains AI Assistant, and copy-pasting code into web-based LLMs (Claude/ChatGPT).
Why I chose Cursor: While GitHub Copilot and extension-based tools are good for single-line autocomplete, they feel bolted onto a standard editor. Cursor was designed from the ground up around AI interaction. Features like Composer (multi-file editing), deep full-codebase indexing, and instant diff reviews make it a true pair-programming environment rather than just an autocomplete assistant. It eliminated the friction of manually switching between browser LLMs and my editor.
Cursor has fundamentally changed the way I build software.
I don't use it to replace engineering decisions—I use it to explore ideas faster, prototype different approaches, and remove repetitive work so I can focus on architecture and product decisions.
What stands out is how naturally AI is integrated into the editor. It feels less like an assistant and more like a collaborative development environment that understands the context of the codebase.
It's one of the few tools that has genuinely improved my daily workflow.
What needs improvement
AI works best when developers stay in control, so I'd like to see even better transparency around why certain solutions are suggested and easier ways to compare alternative implementations before applying changes.
Performance has also improved significantly over time, but very large projects can still feel slower than a standard editor in some situations.
Before switching to Cursor I mainly used VS Code with GitHub Copilot.
Copilot is excellent for code completion, but Cursor provides a much more integrated AI experience. Being able to reason about multiple files, refactor larger pieces of code, and iterate quickly without constantly changing tools made the transition worthwhile.
It fits much better into the way I build products today.
Claude Code
VS Code
Windsurf
Github Copilot