What do you like best about the product?

What I like best about Cursor is that it feels like an actual coding tool built for developers instead of just an AI chatbot inside an editor. The biggest difference for me is the tab completion experience. Unlike CLI agents that generate code somewhat blindly and make broad changes without much visibility, Cursor keeps you in control. You can see suggestions inline as you code, accept them piece by piece, and guide the implementation naturally instead of handing over the entire task.

The contextual autocomplete is incredibly good for real development work. It understands the surrounding files, existing patterns, variable names, and architecture, so the suggestions often feel like something I would have written myself. It saves a huge amount of time on repetitive coding, refactoring, boilerplate generation, and debugging.

I also really like how flexible the platform is. Being able to choose different AI models or connect your own API keys is a huge advantage. Different models work better for different tasks, and Cursor gives you the freedom to optimize for speed, quality, or cost depending on the workflow. That flexibility makes it much more practical than tools that lock you into a single model.

From a performance standpoint, Cursor feels fast and responsive even on larger projects. The inline suggestions appear quickly enough that they don’t interrupt flow, and the editor itself still feels lightweight because it builds on the VS Code ecosystem. I was able to keep my existing extensions and setup without needing to rebuild my workflow.

The onboarding experience was also smooth. Since the interface is familiar, it only took a few hours to fully integrate Cursor into my daily workflow. The documentation and setup around models and API keys are straightforward, especially for developers already comfortable working with AI tools.

In terms of ROI, Cursor has easily paid for itself through time saved. Tasks like writing repetitive code, debugging, refactoring, and understanding unfamiliar codebases are significantly faster now. It reduces a lot of the small friction points that normally slow development down throughout the day.

Overall, Cursor improves productivity without taking control away from the developer. It enhances the coding experience instead of trying to automate everything blindly, and that balance is what makes it genuinely valuable.

What do you dislike about the product?

One thing Cursor could improve is consistency across different AI models and workflows. While the flexibility to choose models and connect your own API keys is a huge advantage, the quality of responses and edits can vary depending on the provider. Sometimes one model handles context perfectly while another loses track of project structure or makes overly broad changes. Smarter defaults and better guidance around model selection would make the experience smoother.

The AI can also occasionally become too aggressive with edits when using larger code actions. In some cases it rewrites more code than necessary or changes files outside the intended scope. It is still better than fully autonomous CLI agents because Cursor keeps the developer in control through inline suggestions and reviewable edits, but the AI can still feel overconfident at times.

Performance is generally very good, but there are occasional stability issues around extensions and AI panels. I have run into situations where extensions like Codex fail to load properly, get stuck on loading screens, or disappear entirely until Cursor is fully restarted. This does not happen constantly, but when it does, it interrupts workflow and can be frustrating during longer coding sessions. There are multiple similar reports from other users around extension loading and chat panel reliability as well. ([GitHub][1])

On larger repositories, indexing and context retrieval can also become inconsistent. Sometimes suggestions lose relevance or the AI misses nearby context that feels like it should have been included automatically.

Pricing can also become a consideration for heavier users, especially when using premium models with API-based billing. The flexibility is great, but costs can increase quickly if you use high-end reasoning models throughout the day. Better usage analytics and optimization suggestions inside the product would help users manage spend more effectively.

From a UI perspective, the core editor experience is excellent because it builds on VS Code, but some AI-related controls and settings still feel scattered across multiple menus. Features like indexing behavior, model routing, permissions, and agent configuration could be organized more clearly.

Support and onboarding are solid overall, but newer users may not immediately understand the workflows that make Cursor truly powerful. The tool becomes significantly better once you learn how to guide the AI properly, combine chat with inline editing, and use different models strategically. More advanced onboarding examples and workflow templates would help shorten that learning curve.

Overall, most of the downsides are related to polish, reliability, and predictability rather than the core product itself. Cursor is already one of the best AI coding environments available, but improving extension stability, large-project consistency, and cost transparency would make it even stronger.

https://github.com/openai/codex/issues/17290

What problems is the product solving and how is that benefiting you?

Before using Cursor, a lot of development time was getting lost to repetitive work, context switching, and debugging cycles. I would constantly move between the IDE, documentation, Stack Overflow, GitHub issues, and separate AI tools just to implement or troubleshoot relatively small things. That workflow was slow and mentally draining, especially when working across larger codebases or unfamiliar frameworks.

Cursor solved a big part of that by bringing high-quality AI assistance directly into the editor in a way that actually fits the development workflow. Instead of copying code into a browser chat and pasting results back manually, I can work directly inside the IDE with inline suggestions, contextual edits, and chat tied to the actual codebase.

The biggest benefit has been speed without losing control. The tab completion system helps generate code incrementally while still letting me guide architecture and implementation decisions myself. Compared to CLI-style agents that can feel blind or overly autonomous, Cursor keeps the developer in the loop. I can accept suggestions line by line, refine prompts based on context, and make targeted edits much faster.

It has also significantly improved onboarding into unfamiliar projects and technologies. Instead of spending hours tracing files manually or searching documentation, I can ask Cursor to explain patterns, summarize modules, or suggest fixes based on surrounding project context. That has reduced the time needed to understand new codebases dramatically.

In terms of measurable impact, routine tasks like refactoring, writing boilerplate, debugging errors, and generating components now take a fraction of the time they used to. For many workflows, it easily saves multiple hours per week. It also reduces cognitive overhead because I can stay focused in one environment instead of constantly switching tools.

Another benefit is flexibility. Being able to choose different AI models or use my own API keys means I can optimize for speed, reasoning quality, or cost depending on the task. That has made the ROI much better compared to tools that lock users into a single model or pricing structure.

From a UI and integration perspective, Cursor fits naturally into existing workflows because it builds on the VS Code ecosystem. I did not have to abandon existing extensions, shortcuts, or development habits to start getting value from it.

Overall, Cursor is solving the problem of fragmented AI-assisted development. Instead of AI being a separate tool outside the coding environment, it becomes part of the actual workflow. That has improved productivity, reduced repetitive work, sped up debugging and learning, and made development feel much more fluid overall.

What do you like best about the product?

It’s easy to use and integrates well with my IDE, but it doesn’t integrate as well with browsers or design docs. Whether I’m using the AI features or not, I still prefer relying on the auto-complete from the base models. My main issue is performance: streaming is very poor, and it takes too long to understand the codebase, which really hurts the overall experience and increases the time it takes to deliver results. On top of that, support from their team is basically non-existent.

What do you dislike about the product?

Token utilization is the worst among all the models I’ve tried. The pricing also feels like it breaks the bank day after day, which makes it hard to justify long-term use. It’s not sustainable for larger projects, especially because it keeps sending recursive context snapshots that quickly drain my token balance. On top of that, the UI needs a complete revamp. In its current, worse-than-imaginable state, it’s basically impossible to use on smaller devices.

What problems is the product solving and how is that benefiting you?

It helps me manage my day-to-day reviews across both my homelab VM infrastructure setup and enterprise cloud product environments, covering everything from C++ to JavaScript. That said, it performs very poorly with low-latency code and anything proprietary—which is expected, since it’s trained the way it is.

What do you like best about the product?

How does it keep the context of the projects I’ve added?

What do you dislike about the product?

For me, the main issue is the limit, because I can reach my monthly limit in only 10 days.
There should be an option for weekly and per session limits

What problems is the product solving and how is that benefiting you?

Reducing time to develop new features

What do you like best about the product?

Cursor is AI powered Integrated Development Environment with access to various frontier models. It helps developers to focus on holistic direction while leaving the implementation pretty straight forward and easy with Cursor.

What do you dislike about the product?

Cursor is based on the VS Code IDE. The default UI feels a bit plain and not very intriguing. Also, sometimes when the context window is about to run out and I keep continuing back-and-forth chats, it can freeze.

What problems is the product solving and how is that benefiting you?

Cursor has become an indispensable tool in my day-to-day work. It helps me across the whole workflow—from writing new code and refactoring existing code to supporting me during code reviews.

What do you like best about the product?

codebase knowledge and grep speed. the ability to run multiple models

What do you dislike about the product?

The pricing is too high. but In the end, is a great product

What problems is the product solving and how is that benefiting you?

let me code in languages i don't know how

What do you like best about the product?

The AI suggestions are actually useful, it understands context pretty well and saves a lot of time when writing code.

What do you dislike about the product?

The free tier is quite limited, have to pay to get the most out of it

What problems is the product solving and how is that benefiting you?

Speeds up the coding process a lot , the AI helps when you're stuck or need to write repetitive code quickly.

What do you like best about the product?

I primarily use Cursor for managing complex, multi-file refactors and navigating large codebases. Using Composer mode to update my backend routes, frontend types, and unit tests in a single prompt has eliminated the grunt work of development. Cursor feels like coding with a senior engineer by my side. The biggest problem it solves for me is the tedious context-switching of modern development by having full codebase awareness, allowing seamless multi-file refactors. I love that the AI is seamlessly integrated into the IDE with features like inline generation and tagging, ensuring high-quality assistance without leaving my editor. Autocomplete feels like it's steps ahead of my actual typing. The inline generation lets me modify code directly with a clear, side-by-side diff view, eliminating the distracting copy-paste routine. The ability to tag specific files ensures the AI gets the exact context it needs, yielding accurate, project-specific code. The initial setup of Cursor was impressive, taking less than two minutes, with zero manual configuration required, thanks to its basis on VS Code which allowed me to import all of my extensions, custom keybindings, and themes in a single click. Cursor has become an indispensable tool, pairing perfectly with design tools like Figma to generate matching React or Tailwind CSS code from UI component screenshots. Transitioning to Cursor from VS Code was smooth, with our custom settings and extensions working instantly, making it a massive upgrade for development speed.

What do you dislike about the product?

While Cursor is a massive productivity leap, it still has some rough edges that could be improved. On massive enterprise codebases, the background indexing can cause noticeable editor lag, and the semantic search occasionally pulls in the wrong file context. Additionally, the multi-file Composer feature requires extreme vigilance during code reviews, as it can easily overwrite custom logic across files if your prompt isn't perfectly narrow. If the team can optimize resource consumption for large repos and fix the aggressive keyboard shortcut overrides, it would be flawless.

What problems is the product solving and how is that benefiting you?

I use Cursor to manage complex, multi-file refactors and navigate large codebases, eliminating tedious context-switching and manual updates. It offers seamless integration of AI in my IDE, improving focus and offering project-specific assistance, making coding feel like working with a senior engineer.

What do you like best about the product?

Easy prompting and quick availability of latest models

What do you dislike about the product?

The UI can be plain and boring. Code reviews are not the best currently and require improvement.

What problems is the product solving and how is that benefiting you?

AI-Assisted software development IDE for enterprises as well as personal use.

What do you like best about the product?

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.

What do you dislike about the product?

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.

What problems is the product solving and how is that benefiting you?

It speeds up development by offering native AI support with multiple models and plugins. For me, it feels like a one-stop shop for developers and automation testers, since everything I need is available in one place.

What do you like best about the product?

I like how Cursor helps me code faster with its AI suggestions and clear, easy-to-follow explanations. It’s also helpful for debugging, and overall it makes my workflow smoother, more productive, and easier to stay focused on.

What do you dislike about the product?

Sometimes the AI suggestions aren’t completely accurate, so I still have to review the code carefully to make sure everything is correct.

What problems is the product solving and how is that benefiting you?

Cursor helps cut down my coding time by offering smart suggestions, catching and fixing errors, and explaining code when I need it. Overall, it makes my development work faster, smoother, and more efficient.