The file context indexing is insanely good. I only use the auto mode, but even in that mode the way it refactors, documents, and generates boilerplate is about 85% of what I’d want. Review collected by and hosted on G2.com.

Cursor is sometimes unresponsive, which I assume is due to high traffic on their servers. It doesn’t happen often, but it’s still kind of annoying when it does.

Also, the IDE’s features sometimes just stop working mid-way. For example, the hover feature that’s supposed to show comments on a function in a formatted UI will sometimes not work at all—hovering shows nothing. Review collected by and hosted on G2.com.

I particularly like Cursor's Agent Mode, which can work across multiple files, understand the entire codebase, and handle more complex development tasks with minimal guidance. The improved background indexing and semantic code search have also made it much easier to navigate large repositories and find relevant code instantly. Another feature I've found valuable is the enhanced support for Model Context Protocol (MCP) and multiple AI models, giving me more flexibility to integrate external tools and choose the model that works best for different coding tasks. Review collected by and hosted on G2.com.

Sometimes the AI suggestions are overly complex for simple test cases, adding unnecessary abstractions. It occasionally generates test code with outdated patterns or deprecated methods from older framework versions. Review collected by and hosted on G2.com.

What I like best about Cursor.sh is how fast and smooth it makes coding. The AI suggestions feel natural and actually useful, especially for debugging and writing repetitive code. It saves a lot of time and helps me focus more on logic instead of boilerplate. Review collected by and hosted on G2.com.

Sometimes the AI suggestions feel inaccurate or a bit too aggressive, especially when I’m working on larger projects. It can also slow down with heavier files, and some of the more advanced features take time to fully understand and use effectively. Review collected by and hosted on G2.com.

What I like most about Cursor⁠� is how seamlessly it brings AI directly into my coding workflow. Features such as intelligent code completion, inline edits, codebase-aware chat, and quick bug fixes save me a lot of development time. Overall, it feels like having an assistant right inside the editor, instead of constantly switching back and forth between different tools. Review collected by and hosted on G2.com.

One thing I dislike about Cursor� is that it can sometimes get overly aggressive with AI-generated changes, especially in larger codebases. We still have to review its suggestions carefully, since it may introduce unnecessary refactors or even incorrect logic. It can also feel resource-heavy at times, and some of the more advanced features seem to depend heavily on paid usage limits. Review collected by and hosted on G2.com.

I’m actively using Cursor in a professional work environment for data engineering and ETL development projects. My work involves writing Spark ETL jobs and debugging large-scale pipelines. I use Cursor for Python development and for troubleshooting these pipelines, and it helps me optimize code, debug Airflow DAGs, and refactor the code I work with most often. Review collected by and hosted on G2.com.

I don’t have any major dislikes about Cursor; it’s a good product overall. Review collected by and hosted on G2.com.

The AI-powered autocomplete is genuinely impressive — it understands context across multiple files, not just the current one. The inline chat and CMD+K for quick edits make it feel like pair programming rather than just code completion. Tab completions are fast and accurate enough that I actually trust them. Review collected by and hosted on G2.com.

The support experience leaves a lot to be desired — response times are slow and issues often go unresolved for days. The billing and subscription management can be confusing, and when things break (like context window limits or sync issues), there's not much transparency on what's happening. The documentation also lags behind new features. Review collected by and hosted on G2.com.

The UI of the agents window, and how I can see the changes directly in the code and in the UI field. The speed is also outstanding, and the choice is great. You offer all the flagship models at reasonable prices. We pay just 60$ and the value is amazing. We also integrated it into many parts of the system, and it's great. AI is clever, and the program's onboarding is really simple; you just code. Review collected by and hosted on G2.com.

Perhaps that can remove older models of the Composer, i really like to use them for the simplest tasks to save costs, even though the cost is already low :D Review collected by and hosted on G2.com.

I like that Cursor is user-friendly, which makes it easy to navigate. The main thing I like about Cursor is that it feels intuitive. I also appreciate the picture-in-picture mode, which works really well. I find the code base indexing in Cursor very helpful because it allows the AI to understand the project structure and provide context to suggestions. It reduces my cognitive load and speeds up the transition from an idea to working code. Cursor enhances my ability by understanding the relationship between different files and significantly reduces context switching, which boosts my productivity. The initial setup of Cursor was easy because I could install it like a normal desktop app and launch it from my existing VS Code settings, extensions, and themes. Review collected by and hosted on G2.com.

I wish Cursor was more like the Chrome browser. I use it mainly for software development, but I'd like to use it for everyday tasks too. If its AI native features could automate the creation of repetitive tasks and the development of integrated systems, it would be more useful.The pricing of the software could have been better against performance. Review collected by and hosted on G2.com.

What I like most about Cursor is that it does understand the context of the entire project, not just the line where you are.

For example, when working with FastAPI, I can ask it to adjust an endpoint and it automatically takes into account models, services, and even other related files. This saves a lot of time compared to copying and pasting code into a chat.

It is also very useful for refactors: you tell it "change this logic throughout the project" and it does it quite well, maintaining consistency. This is really felt in real projects.

And a key thing is that it serves more for thinking than Copilot. It not only autocompletes, but you can discuss a solution directly on your code, as if it were another dev deeply involved in the repo. Review collected by and hosted on G2.com.

What I don't like is that sometimes it goes overboard with being "creative."

When you ask for something specific, it can end up changing more than necessary or proposing a solution that doesn't quite fit with your architecture. In projects with already defined logic (like well-structured APIs in FastAPI), this forces you to review everything with a fine-tooth comb.

It also depends a lot on the context it has loaded. If it doesn't fully understand the complete flow, it starts to assume things and that can break coherence between files. Review collected by and hosted on G2.com.

I use Cursor on a daily basis as my best buddy. It helps me write code, create AI agents, debug issues, find the RCA, do POCs, and—most importantly—create documents for my team and org about the work I’m doing. Review collected by and hosted on G2.com.

Sometimes, when working with multiple agents, it uses too much memory. Apart from that, there’s nothing else to mention. Review collected by and hosted on G2.com.