Cursor Reviews & Product Details

What I like most about Cursor is its IDE integration and lightweight performance. It has a better system design and is faster in creating UI designs. The inline text edit feature is also great because it allows me to make changes efficiently, like converting blocks of code into async functions or adding error handling without rewriting everything. This feature lets me quickly generate components compared to other software. The integration with my editor reduces the need for context switching between tabs, which really smooths out my coding sessions. Review collected by and hosted on G2.com.

It was good back in 02/1424, but right now in 2026, it has bad usage limits and poor model reasoning. Also, it has bad ROI as I'm spending my money and not getting the ROI I should. Considering a cloud code or OLEX subscription is much better than this. Review collected by and hosted on G2.com.

I really like Cursor's AI code generation because it automatically generates all the code for me, including complete functions, classes, and entire files without me having to type everything manually. It understands my project structure, which is top-notch. The detailed step-by-step instructions for setting up development and production environments save a lot of manual steps, especially for junior developers. It has significantly reduced the time for MVP creation from weeks to days and eliminated the need to write boilerplate code manually. Cursor is also great for generating 70-80% of the MVP code without manual typing, handling tasks like database schema, API endpoints, and authentication. This has saved us about twenty to thirty hours per week per person on routine coding tasks. Cursor generated the entire project structure, including database connections, API routes, and frontend components, in just one session. It makes our jobs easier by providing clear instructions on installing databases, APIs, and dependencies, significantly reducing the learning curve. Review collected by and hosted on G2.com.

While Cursor is excellent, there are some areas that could be improved. The AI becomes less helpful for fine-tuning complex customizations. It generates great initial code but struggles with advanced features, making more errors. Its performance can be inconsistent, sometimes perfect, other times producing broken or insufficient results. This inconsistency makes it unreliable for critical production issues. Additionally, Cursor needs to improve context retention to avoid breaking what’s already built, as it often makes errors that need constant context clarification to prevent mistakes. It still requires human oversight and assistance for debugging. Review collected by and hosted on G2.com.

AI-native experience: It's not just a plugin bolted onto an existing editor; the AI is baked into the whole workflow, so suggestions and edits feel natural rather than like a separate tool you're switching between.

Codebase awareness: It understands the context of your whole project, not just the file you're in, which makes suggestions far more relevant and reduces the back-and-forth of explaining what you're working on.

Tab completion that actually thinks: The autocomplete goes beyond finishing a line; it anticipates multi-line changes and whole block rewrites, which saves a lot of time on repetitive edits.

Chat with your code: Being able to highlight a section and ask questions or request changes in natural language, without leaving the editor, is genuinely useful rather than just a gimmick. Review collected by and hosted on G2.com.

Can be overconfident: It sometimes generates code that looks right but has subtle bugs; you still need to review everything carefully, which can create a false sense of security if you're moving fast.

Suggestions can get in the way: When you know exactly what you want to write, the constant AI suggestions can feel interruptive rather than helpful; turning them off defeats the purpose a bit.

Context window limits: On larger codebases it sometimes loses track of earlier context, so suggestions become less relevant the more complex the project gets. Review collected by and hosted on G2.com.

The parallel sub-agent and the agentic flow are really nice. Review collected by and hosted on G2.com.

I don’t like the way the Cursor team changes the token limit without letting users know; that’s really bad. It also doesn’t allow me to start a new chat while keeping all the context from the previous chat, which slows down my agentic workflow because I have to explain the context to the agent all over again. Review collected by and hosted on G2.com.

- UX/UI is very simple and intuitive;

- Many integrations to the other tools and libraries;

- Fast and barely any downtime;

- About 10x ROI in my experience;

- Haven't worked with support yet, but it's just because I didn't need it yet.. Cursor is just working. My first time/onboarding was a bit slow, but that was on an old UI. Now it's much better.

- Newest AI models are available and it's awesome. Plus sometimes there are some deals ;) Review collected by and hosted on G2.com.

I think one downside that I saw so far was rare downtime of an agent. But that might be that it's just a model that I was using. Review collected by and hosted on G2.com.

It includes integrated AI as well as a separate, agent-only mode. Depending on your flexibility needs or specific requirements, you can switch between them. It also supports nearly every frontier AI model, which is really great. Review collected by and hosted on G2.com.

The cost can sometimes be unpredictable. Even though there’s a tracking dashboard, we still often end up surprised by our AI expenses. Review collected by and hosted on G2.com.

Cursor is able to generate good-quality code that matches our organization’s coding practices. We’ve been able to reduce the overall development cycle significantly. It’s also helpful for finding bugs and resolving them quickly. We were also able to reduce our tech backlog, as most of the mundane coding tasks were easily done by cursor. Review collected by and hosted on G2.com.

Support is currently limited to a small set of models, and I’d like to see newer models integrated as well. Also, when I run multiple chats at the same time, the IDE becomes noticeably slow. Review collected by and hosted on G2.com.

One of the things I love most about Cursor is its ability to access the whole repo. This feature is extremely empowering, especially for someone like me who doesn't come from a traditional coding background. It helps me understand and trace the logic we've implemented, which is incredibly helpful during the alpha stage of prototyping. Cursor's capability to refactor and rename things across multiple files is huge for my workflow. It also enhances debugging by allowing me to keep iterating and finding solutions in real time. Additionally, I love how Cursor can show me the code like a more classical coding tool, helping me grow into understanding and taking control of my code. Its ability to help prototype bits and prove workflows possible is also amazing. Overall, Cursor is extremely helpful for my development in an entrepreneurial sense within the company. Review collected by and hosted on G2.com.

I think one of the biggest challenges with Cursor is that it can lead to code that becomes really bloated and messy, what the other developer called 'spaghetti code.' We had to bring another developer to fix and make the code more maintainable. The advice I got about whether to keep things in one extension or multiple scripts led to a large amount of bloat. It would be helpful to have more human-based guide rails for maintainability, suggesting best practices for setting up projects. Without them, the AI might keep developing something that doesn't make sense and becomes unmaintainable. While Cursor can rush to a prototype, having guidance early in the process on these practices would be beneficial. Review collected by and hosted on G2.com.

I really love Cursor for its powerful AI assisted coding, especially how it can understand my codebase and generate relevant code suggestions or edits instantly. In my daily work, it saves me a lot of time by helping me with debugging, writing the boilerplate code, and even explaining the complex logic step-by-step in a simple way. The UI feels clean and familiar (like the VS Code), which made it easy for me to get started without a steep learning curve while still boosting my productivity significantly Review collected by and hosted on G2.com.

I don't have any reason to dislike Cursor, but I sometimes find Cursor’s AI responses inconsistent, especially with more complex tasks, which means I still need to verify and refine the output sometimes. In my experience, performance can slow down when working on larger codebases, which affects the overall flow. I also feel the pricing could be more flexible Review collected by and hosted on G2.com.

The interface feels similar to popular IDEs, so as a developer I found it easy to pick up and use. It offers multiple LLM models to choose from, which is helpful depending on what I’m working on. The price point is on the high side, but it still feels worth it for what you get. I also like that it can integrate with third-party MCPs, making it possible to build agentic workflows. Review collected by and hosted on G2.com.

The reported model unavailability doesn’t align with the models’ actual uptime. Review collected by and hosted on G2.com.