I use Claude for context awareness and structured reasoning, particularly because it handles large multifile projects better than simple code completion tools. I find it helpful for reviewing and debugging code, as it provides deep explanations and reliable step-by-step suggestions. Claude efficiently solves various issues with accurate solutions, making it a valuable tool for my programming tasks. It's especially beneficial for debugging, refactoring, and understanding multiple aspects when handling single-line commands, and it’s a lightweight tool for tiny scripts. The setup process is easy and straightforward, just plug and play. Review collected by and hosted on G2.com.

There are some limitations and drawbacks with Claude. It sometimes becomes inefficient, especially during long sessions, and doesn't always provide accurate results if I'm continuously querying it for over an hour. It can lead to false file setups and delay projects, creating messy content changes. There's also a learning curve when it comes to editing and efficiency tools in longer sessions. Review collected by and hosted on G2.com.

Claude has a way of writing AI in a very human like manner, which is better than some of the other AI tools. It does a great job of reasoning and thinking about the next steps to do, even without being asked. Claude has the ability to leverage Projects, Skills, and Cowork, which are all advanced features which are good to use and help save time. Projects are a particularly helpful way to organize your chats and separate themes inside of your AI usage. Review collected by and hosted on G2.com.

The only downside of Claude is that it throttles your usage when you hit limits for the $20 a month option, which is quite often if you are using it heavily. Also, it depends on your organization on whether or not you'll be able to give it access to business documents and integrations with other work software. Review collected by and hosted on G2.com.

I like how natural and thoughtful the writing feels with Claude — it actually sounds like a person, not a stiff template, and it picks up on tone really well. It's good at following nuance and admits when it's unsure instead of just making things up. I appreciate that the long conversations hold context nicely, so I don't have to keep re-explaining myself, which is especially helpful when working on bigger projects. Claude feels like an ongoing collaborator, helping me communicate more clearly, and saving time by remembering details and the direction the conversation is heading. Also, the setup was super easy—just signed up and started chatting with no real setup or learning curve. I was up and running in a couple of minutes. Review collected by and hosted on G2.com.

A few things. It can be a bit wordy and over-explain when I just want a quick answer, so I sometimes have to tell it to be concise. It's also cautious and will occasionally hedge or add disclaimers I don't need. And since it doesn't always have the latest info, I have to double-check anything time-sensitive. Better handling of very large files and more up-to-date knowledge would make it even stronger. Review collected by and hosted on G2.com.

It is incredibly good with coding. Especially Claude Sonnet-4.6. It could easily create readily launchable HTML based websites with over 1500 lines of code in a single descriptive prompt. Review collected by and hosted on G2.com.

Upon creating code files containing 1000+ lines, the chat often becomes unresponsive. The AI struggles to respond. Also, any further improvement prompts leads to crash. Also, it overkills by unnecessarily visualizing the data when we are simply trying to understand without visualizations or when we are simply trying to filter it. Review collected by and hosted on G2.com.

Claude stands out for its ability to handle both technical and creative tasks with equal depth. Whether I'm debugging code, drafting documents, or brainstorming research ideas, the responses are thoughtful and well-structured rather than generic. The conversational memory within a session makes complex multi-step workflows feel natural. It's also noticeably honest — it flags uncertainty instead of confidently hallucinating, which matters a lot when you're using it for serious work. The interface is clean, and the model genuinely feels like a capable thinking partner rather than just a search engine replacement. Review collected by and hosted on G2.com.

The context window, while large, can still be a bottleneck during very long sessions — earlier details sometimes get deprioritized in extended conversations. The free tier's message limits can interrupt flow during intensive work. Occasionally, Claude can be overly cautious on technical security topics, adding unnecessary caveats even for clearly legitimate use cases. There's also no persistent memory by default across sessions, so you have to re-establish context each time. A native code execution environment for testing snippets directly would be a welcome addition. Review collected by and hosted on G2.com.

What I like most about Claude is how clearly it explains things and how well it structures its answers. It also seems to handle more complex questions better than most tools I’ve tried. It’s especially helpful for writing and organizing ideas, although at times the responses can come across as a bit generic. Review collected by and hosted on G2.com.

What I dislike about Claude is that its answers can sometimes feel too generic or overly cautious. It also tends to produce unnecessary documentation or long-winded explanations when I’m really just looking for a simple, direct solution. On top of that, it can be a bit repetitive at times, which makes the responses feel longer than they need to be. Review collected by and hosted on G2.com.

I use Claude for my development work and daily tasks since I'm a game developer. It solves a lot of problems for me, especially bug fixes and architecting my software in a better way. I'm able to find corner cases and try to solve bugs without going into a rabbit hole of code. It gives me a lot of context and easy problem solving, and very concise answers to complex problems. What I like most is that Claude is very accessible and very concise, and it gives very straightforward answers without hallucinations. When I was spending a lot of time trying to fix a bug where the UI was not showing up, I gave Claude a bit of context about which file it should look for and it gave me an exact result. It also helped me drill down problems when I was working on optimization, giving me solutions to resolve memory leaks and reduce CPU usage based on the code architecture. Claude has access to my project code, so it was able to go through the whole code, get enough information about where the code is linked, and resolve those issues very easily. The Claude UI is very accessible, and you can ask non-development problems as well. Like filing taxes, which I recently did, and it helped me a lot with that and gave me very straightforward answers with comparisons and everything that I needed to hear. The initial setup was quite easy, we just had to install the Windows version and set up the login process and you're all set. My team prefers Claude much better because it's more concise and straightforward and it was the tool we needed for development purposes. We had the enterprise version of Claude, which was far better in terms of answering questions. Review collected by and hosted on G2.com.

I'd like Claude to improve their tokenization. When I start a project, I have to give a lot of context, and that uses up a lot of tokens. If those useless tokens could be compressed, the API billing usage would be far lesser than what it is today. Also, Claude can't do image generation, so for that I still have to use ChatGPT. Review collected by and hosted on G2.com.

Even when I use the lowest-capacity model, the reasoning is still complex and it gives accurate answers every time. It doesn’t waste time flattering you; it gets straight to the point and executes. It also doesn’t start working with little or no context—it asks for clear instructions before proceeding. So far, it’s the only model that has done that for me.

It understood my business in a few lines of context and it's understanding is growing as we work in different scenarios/cases Review collected by and hosted on G2.com.

It's frustrating when it run out of tokens even when we are working only on simple tasks that doesn't require complex reasoning Review collected by and hosted on G2.com.

Claude code understands requirements like system design very well and can produce production-level code in a single pass. One useful feature is that it can switch to a compact response when tokens are about to run out, so the output still stays complete and readable. Review collected by and hosted on G2.com.

It takes longer to generate responses. The visualization and graphical feature-building also feel more complicated to use, and token consumption seems higher compared with other options. It's struggles with conceptual work for understanding. Review collected by and hosted on G2.com.

There are a couple of things that I really love:

1) The model is really powerful and I have to say that Opus is the best way to go for coding

2) I love how easy to use and powerful for MCPs, Skills and in general all the integrations Review collected by and hosted on G2.com.

Sometimes it feels really dumb, especially when it gets a lot of context. Also, switching between sessions or remembering things across sessions isn’t that good. I would really like an easier subscription/account change option, since I use both a work and a personal subscription and want to be able to switch easily between those two accounts. Review collected by and hosted on G2.com.