The best part is the interactive interface, with support for multiple LLM models and an agent mode that lets me hand off tasks directly to the LLM. I also really like being able to create workspaces with multiple repos, along with the integrations with various tools. Review collected by and hosted on G2.com.

To be honest, I don’t have anything major to complain about. That said, sometimes it takes a while to respond to queries and apply changes to files. It might be related to the context window, but I’ve noticed occasional performance slowdowns from time to time. Review collected by and hosted on G2.com.

Cursor is a very powerful AI-assisted code editor that significantly speeds up development. The AI integration feels natural and is deeply embedded into the workflow, making it easy to generate code, refactor functions, or understand unfamiliar parts of a codebase. It’s especially useful for navigating large projects, where you can quickly ask questions about the code and get relevant context-aware answers.

The interface is clean and similar to Visual Studio Code, so onboarding is quick. Features like inline suggestions, chat-based assistance, and the ability to modify multiple files at once make it very efficient for day-to-day development. Overall, it helps reduce repetitive work and improves productivity. Review collected by and hosted on G2.com.

While the AI features are very helpful, they are not always perfectly accurate and still require validation. For complex or critical logic, you need to carefully review the generated code. Performance can also vary depending on project size and usage. Additionally, relying heavily on AI suggestions may reduce deeper understanding if not used carefully. Review collected by and hosted on G2.com.

The agentic workflow it enables, along with the clean UX/UI and the plugins they’ve added recently, is honestly a lifesaver. Having it in the cloud and watching it deliver a full end-to-end test demo is kind of insane.I use this as it's subagents and everything is just out of league for any other IDE. Review collected by and hosted on G2.com.

Pricing is so absurd. I’m a pro user, and after just a few prompts my monthly quota is already gone. Given how much development I do, I have to be on auto for the whole month. I certainly love all the features it does have all the features than any agentic IDE of nowadays. Pricing should be transparent tho. Review collected by and hosted on G2.com.

I mostly use Cursor for coding. It helps me with developing the architecture of new projects, handling day-to-day development tasks, making critical decisions, refactoring old codebases, and enhancing UI/UX interfaces easily. I really appreciate the live web inspector, plan mode, and the ability to run multiple agents in parallel. These features make me work faster on my projects and save time. I find that Cursor offers better LLM responses and handling of codebases, providing a better IDE for more quality work, and it operates faster. The initial setup was very easy. Review collected by and hosted on G2.com.

Running terminal tasks in Windows using PowerShell needs improvement. It always gets stuck, and I have to stop the process. It would be better if this was running smoothly and fast as in macOS, thus making my work faster. Review collected by and hosted on G2.com.

I like that I can have full control over the chat responses and output, which really helps me customize my experience. Adding a /continual-learning loop is fantastic as it makes a learning system that keeps improving by evaluating the process and optimizing for better, faster, and more effective results next time. Even though I don't know much about coding, Cursor allows me to request changes to GUI settings and logs them, which is super helpful. The fact that it took a while to get everything set up just right and now it works 90% of the time right away is a testament to its capabilities. As my first GPT-IDE, after a lot of work, Cursor has really helped me take off. Review collected by and hosted on G2.com.

I still have to battle a lot of assumptive thinking, where the model fills in nonsense or details that I didn't request, or sometimes decision making goes through some nonsense steps and wastes tokens. Also, it's taken well over a year to get to a point where I feel confident and comfortable that the model will get about 90% right the first time. Review collected by and hosted on G2.com.

It is a new way of programming. It helps when I need it but does not come pushy with proposing changes. The UI is old school, but I like it this way. I've been suing Visual Studio before I found them pretty similar. I was able to download my old setup so I did not need to configure it all over again. Performance is great - I get responses really fast. I like the Composer 2 (AI model) feedback on multiple files (to be able to comprehend the full project). Review collected by and hosted on G2.com.

To be honest I did not find anything that I would not like. Composer 2 AI model is quite expensive but compared to Auto (which is usually Claude or OpenAI) it really shows the value. I did not need any help with the setup -as well. Review collected by and hosted on G2.com.

Cursor is the best agentic AI assistant for full-stack development. It handles back-end and front-end tasks, including n8n workflow optimization, with incredible ease.

Efficiency: Performance is so high that I no longer need a team of 2+ people.

Integration:Seamless workflow with VPS deployments and Supabase.

Features: The Browser mode is a killer feature for instantly identifying and fixing problems.

Value: Unbeatable pricing for the power it delivers. Review collected by and hosted on G2.com.

When I work with n8n workflows (*.json scheme), the application sometimes crashes. Also, creating new flows with Cursor is tricky: I first have to create templates, then point it to what needs changing, and only after that does it work. Review collected by and hosted on G2.com.

The interface and user experience are excellent.

It perfectly adapts to a developer's workflow.

I really liked being able to clearly see in the interface when the agent is editing the code, which allows me to monitor the changes accurately.

Also, something I'm still experimenting with, but I managed to connect my own MCP server with Cursor.

The implementation process was perfect because I was used to using Visual Studio Code as an IDE, so it was very quick and smooth to implement Cursor in my work as the main IDE and recommend it to the rest of the team I am in charge of.

I also really like the automations I achieve with the commands. Review collected by and hosted on G2.com.

It is still very expensive to use the LLM AI models that work best. And it's difficult to achieve a return on investment. Review collected by and hosted on G2.com.

cursor has been a huge time saver for my automation and API testing work. It helps me generate test scripts, debug failures, and add exception handling faster. AI understands project context well, making it easier to work across multiple files and large automation frameworks. I also like the clean UI and AI agent mode, which reduces manual effort on repetitive tasks Review collected by and hosted on G2.com.

Sometimes Cursor makes changes across multiple files that are not exactly what I intended, so I still need to carefully review the generated code afterward to make sure everything matches what I meant. Review collected by and hosted on G2.com.

I use Cursor as my primary AI-powered code editor for software development tasks, and I really appreciate how it helps with writing new code, debugging issues, understanding existing codebases, and more, especially for languages like Java, Spring Boot, JavaScript, React, SQL, and API development. The deep AI integration in the development environment is something I like a lot, as it understands the context of the entire codebase to provide relevant code suggestions, explanations, and debugging assistance, which boosts productivity. The chat interface is also a favorite feature because it makes it easy to ask questions, generate implementations, and refactor existing logic without constantly switching between different tools. Having this kind of integration directly in the editor saves a lot of time by generating boilerplate code, suggesting improvements, and providing troubleshooting assistance during debugging. I also find it valuable that everything is integrated into the code editor, allowing me to ask questions, generate code, review changes, and refine implementations without switching between documentation sites and search engines. This really streamlines the workflow, allowing me to focus on solving business problems and improving code quality. The seamless integration of AI-assisted capabilities makes daily development tasks much more efficient. Review collected by and hosted on G2.com.

While Cursor is a powerful development tool, there are a few areas that could be improved. Occasionally, the AI may generate solutions that appear correct but require additional validation and testing before being used in production. For complex business logic or large-scale architectural decisions, manual review is still necessary to ensure accuracy and alignment with project requirements. Response quality can also vary depending on the complexity of the codebase and the context provided. In some cases, more precise understanding of project-specific conventions would help improve suggestions. Additionally, AI-generated responses can sometimes be slower when working with large repositories. Enhancements in performance, context awareness, and consistency of suggestions would make the overall experience even better. Review collected by and hosted on G2.com.