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?
@OpenAI is winding down its contract with @Cursor, following Cursor's acquisition by SpaceX.
The two companies have worked together for nearly four years. But OpenAI's contract with Cursor included a clause allowing cancellation after a change of ownership, and they're using it. They have all the right to do so.
Reviewers mostly see Cursor as a strong AI-first editor that fits easily into existing workflows, especially for VS Code users, with standout strengths in codebase awareness, multi-file edits, debugging, refactoring, and fast inline suggestions. Makers of Product Hunt and viktor.com say it works best when developers review the output closely, while makers of Basedash: AI data analyst credit it with faster shipping. The main complaints are heavy resource use, occasional context loss or hallucinations on large repos, frequent updates, and confusing or costly usage limits.
The best coding agent for a while. it is expensive though. you pay a premium price for a premium agent.
What needs improvement
subscription cost (9)
pricing is an issue, especially because that different plans that everybody has. there is not enough clarity about pricing. one can spend 2 requests (same for any other model) for an opus 4.6 request, while other can spend 7 dollars
Cursor is just better at using the given model up to its limits. None of the other agentic products success that much. Also plan mode is a beast if you give some time to your task and plan everything carefully. While coding with the popular tech stacks, Cursor executes the plan flawlessly most of the time.
Cursor has completely revolutionized my development workflow. Since it's built on top of VS Code, the transition was completely seamless, but the productivity boost is insane.
The codebase indexing is magic—the AI actually understands the entire project, not just the file I’m looking at. Features like 'Composer' and smart inline edits feel like having a senior developer pair-programming with me 24/7. It doesn't just speed up my coding; it makes debugging and refactoring effortless. Highly recommended to any developer!
What needs improvement
A couple of minor pain points for power users:
Large Mono-repo Handling: When using 'Composer' to refactor code across dozens of files in a massive repository, it can occasionally miss context or hallucinate minor details.
Release Stability: Because the team ships new features so fast, occasional minor bugs pop up in new versions (though they are usually hotfixed within hours).
I'd love to see more fine-grained controls over which files get indexed to save memory.
I was a long-time user of GitHub Copilot. While Copilot is great as an enhanced autocomplete tool, Cursor operates on a completely different level.
Copilot feels like a plugin slapped onto an existing editor, whereas Cursor is built from the ground up for AI. The ability to reference whole repositories using @ and generate changes across multiple files simultaneously made Cursor the clear winner for me.
This product is a good solution when you start doing software. But first I think you should to learn some lessons from Cursor docs. Otherwise you could burn all your dollars on first iterations of working with it.
What needs improvement
high memory usage (3)subscription cost (9)thinking-loop (1)
As I told before you could burn your token in during of the day. And when you working by Auto regime, it could send your request to gemini fast and you see the same loop of thinking which does not stop and going to burn your tokens. Therefore it's better to test some models by self before come to Cursor. And yes, I understand that Cursor can't to share the same level of tokens as other AI providers. But anyway even 20/60$ subs does not align to my consume and 200$ is expensive for me.
vs Alternatives
Because I could work seamless with my servers, do bugfix, I really love bugfix regime how it works and debug by self, without my participation.
Codex by OpenAI
Claude Code
Antigravity IDE for Vibe Coding
Github Copilot