I switched from Gemini to Claude and it’s a mixed bag

I love Claude. But I also hate it. That is the simplest way to describe my experience after making the switch from Google Gemini to Anthropic’s AI.

I’ve been using Claude daily for over a month now. While the initial honeymoon phase was filled with impressive and authentic conversations, the novelty is wearing off, and I’m starting to miss Gemini. But switching back completely doesn’t feel right either. Here’s why.

Authenticity over flattery

Talking to Claude feels fundamentally different than talking to Gemini. It feels more human, primarily because Claude isn’t afraid to push back.

It has no problem telling me I’m wrong and is constantly challenging my way of thinking. While Gemini generally does this as well, Claude takes it to a more realistic level. It gives me the vibe that I’m chatting with a professional in a specific field rather than an assistant trying to flatter me.

There is also significantly less fluff. Gemini has a tendency to over-praise every prompt, often starting responses with, “That’s the best possible question you could have asked,” even if the query was basic. Claude gets to the point faster, saving me the eye-rolls ten times a day.

One area where Claude objectively wins is the Projects feature. This is Anthropic’s answer to Gemini’s Gems, and it is simply a better implementation.

In Claude, I can organize conversations into specific Projects. For example, I have a Fitness project where I track my workout plans, diet, and training issues. Every chat is tucked neatly away, making it easy to find specific info later.

Gemini, by contrast, still relies on a massive, chronological list of chats in the left-hand menu. Once you hit a few dozen conversations, finding that one specific prompt from three weeks ago becomes a chore. Claude also includes a “Project Knowledge” section where I can upload specific docs I want the AI to always remember within that context — a feature that feels much more cohesive than Google’s current offering.

Every AI has limits, but Gemini’s approach here is truly frustrating. On the Gemini paid plan, you never truly know how close you are to the limit until you hit a wall and are told to come back tomorrow.

Claude’s system is much more transparent. It provides a visual indicator of your usage within the current rolling window. Seeing that status bar helps me pace my work so I don’t get cut off in the middle of a critical task.

I am really impressed with the Artifacts feature. Claude was able to create an app based on my prompts that worked perfectly. I tried a few different ones, from to-do apps that mimic my favorite ones I’ve used on Android, to money management apps that keep me on top of my finances.

It’s incredible how fast and good these apps turn out. What would have required weeks of coding in the past can be done in mere minutes with Claude, with simple text prompts. Granted, you have to be quite precise with your prompts to avoid issues, but creating apps has become extremely easy with Claude.

Sure, there are other steps if I actually want to use this outside of Claude and put it on the Play Store, and I definitely haven’t gone that far, but it’s still very impressive what it can do in a short period of time. The AI is known as the best coder, beating out Gemini and the rest of its rivals.

While better in many ways, Claude is way behind Gemini in several areas. The first issue I noticed is how fast I reach Claude’s limits. I don’t think I ever hit a limit on Gemini’s paid plan, while I was able to hit the weekly Claude limit within just three days. Considering its $20 per month subscription, I feel like the limits should be increased substantially.

Then there are all the extra features. Image generation is the one I miss the most. Not only can Gemini create realistic-looking images, but it can also edit them to perfection. And Veo 3 can create short videos, showing off what AI can do.

Claude doesn’t have any image or video generation features at all, which is why I had to switch back to Gemini after the first week just for the images alone. I also like Gemini’s Guided Learning feature that asks questions, forcing me to really think about a problem or topic I’m working on. Not everyone is a fan, but I am, and it’s something I’ve missed ever since moving to Claude.

Claude also has issues accessing specific websites I link to. For example, I added a link to a piece of clothing I bought, asking if I can put it in the tumble dryer without shrinking it. In a lot of these cases, Claude wasn’t able to access the page and see what was on it for whatever reason, while Gemini had no such issues at all.

Then there’s the voice mode, which is where Gemini is far superior and less buggy. I found it sad and entertaining that while speaking to Claude on desktop, the mic on my desk picked up Claude’s response, and the AI thought it was me talking when it was in fact hearing its own voice. It then started talking to itself, which went on forever. It is worth noting that I haven’t experienced this issue in the last week, so it looks like Anthropic fixed it.

Additionally, because of the switch to Claude, I also lost access to various integrations with Google services, such as YouTube Music.

Neither of them. While they seem very similar at first glance, there are more than enough differences between them to make the whole experience completely different. I prefer Claude’s Projects to Gemini’s Gems, and I just generally prefer chatting with Claude over Gemini on most topics. With that in mind, I’m sticking with Claude for the majority of my usage.

However, I won’t be ditching Gemini completely. It’s the AI I turn to as soon as I hit my limits on Claude. I also use it for creating images. For example, I’m in the process of redoing my wardrobe, and since I’m not that fashion-oriented, I have a hard time figuring out what looks good or bad on me. Gemini was able to make images of me wearing different styles and color combos, and helped me find my style alongside a color palette. It’s something Claude couldn’t really help me with, since it doesn’t have proper image-generation capabilities.

Chatbots are tools that help us get stuff done, and each one of them is different. A different job needs a different tool, so instead of proclaiming one chatbot as the winner, I’ll stick to using both. What about you? Let me know in the comments.