Grok 3 Tested: I Love It, But I’m Sticking With Gemini For Now

Ryan Haines / Android Authority

The new Grok 3 was launched a couple of weeks ago, and I’ve been playing with it ever since. I switched to it from Gemini, just to see how the two compare. And since Grok is currently only available to X Premium subscribers, I compared it with the Advanced version of Gemini that’s part of Google’s One AI Premium plan.

Grok 3 is said to excel at various things including coding, but I tested it out for my specific use case. I use chatbots to help me with work by finding typos in my writing, suggesting better headlines, and more. I also use them to analyze files and do research on all sorts of topics. So that’s going to be the focus of this comparison.

Grok 3: How good is it?

I tested both Grok 3 and Gemini side by side with the exact same prompts to see which one is better. I used two Gemini models — 2.0 Pro Experimental and 2.0 Flash — since they have different features. The Pro, for example, is the superior model on paper, but it doesn’t support file analysis, which is why I had to use the Flash model for that particular task.

I started off with a simple task of asking both chatbots to find typos in a long block of text. There were plenty of those inside, and while both did a decent job, Gemini pulled ahead. It found 72 typos while Grok highlighted 66. Grok missed a few basic ones, although Gemini also missed a few that Grok picked up, so neither of them did a perfect job here.

I also asked both chatbots to come up with a few headline ideas for a post, and they both did an excellent job. I then selected my favorite one and asked for five variations of it, and I was impressed with the results both Grok and Gemini delivered. However, this was to be expected, and I didn’t anticipate seeing major differences between them with such a simple task.

Grok was much better at providing a stock market summary.

I then stepped it up a notch, and here’s where I saw the first difference between the two. When asked for a stock market summary that should include the top 10 companies that gained and lost the most in a specific trading day, Grok was way better. It provided me with a quick market summary followed by a list of the top 10 gainers and losers, along with the percentage changes and what caused them. The data provided was clear and concise.

When I asked Gemini the same thing, I got a rather strange response. The chatbot said it could not provide market data for the future and just gave me a hypothetical summary for demonstration purposes, as you can see in the screenshot above. I found absolutely no value in that. I then switched to the Flash model and asked the same question, and to my surprise, the answer was far better. Gemini provided me with a quick market summary in bullets and shared a list of four of the top gainers and losers. It’s better than nothing, but I asked for 10. What’s worse, it didn’t even provide the gains and losses in percentages, which is the data I’m interested in.

Mitja Rutnik / Android Authority

When asked about the number of apartments in the new complex, Gemini just gave me a one-sentence answer with the number, which was correct. Grok, on the other hand, went the extra mile and provided the number of apartments in each building that make up the new complex, alongside the apartment numbers for each building.

It’s worth mentioning that Grok’s longer and more detailed responses aren’t always a good thing, as it’s a matter of personal preference. So, while it does go into more detail, Gemini generally gives me the answers I’m looking for, but it sticks to the bare minimum. What’s better comes down to personal preference. While I generally like a longer response, especially on the topics I tested Grok 3 on, shorter ones are also suitable when nothing more than the basics is needed.

It wasn’t a clear victory for Grok, though. The chatbot didn’t find the answer to the question regarding the flooring installed in the apartments, saying that the document doesn’t contain that info, even though it does. Gemini was able to find it, though.

Grok also has the upper hand when it comes to questions about politics. Gemini doesn’t want to talk about this topic in most cases and refuses to answer questions like, “Who is the president of the US, and is he any good?” while Grok had no issue with that.

Grok vs. Gemini: Which is the better chatbot?

Robert Triggs / Android Authority

Grok 3 is great. It doesn’t get as much attention as Gemini, ChatGPT, or even the new DeepSeek, but it’s just as good. Better in some cases. Like my testing showed, it’s better than Gemini when it comes to researching topics like the stock market as well as when analyzing files. It’s also not as censored, so you can talk politics and other touchy subjects with it.

However, Gemini still has plenty of advantages over Grok 3, so it’s hard to proclaim one as the overall winner. For example, Gemini’s Imagen 3 image generator is better than Grok’s, even though the latter is more fun as it has fewer limitations.

Gemini also has Gems that let me customize the responses I get, Gemini Live that lets me have a real conversation that speeds things up, and extensions that work with Gmail, Maps, and other Google services. A Gemini Advanced subscription also includes access to NotebookLM Plus, which is one of the best AI tools out there.

So, for now, I’m sticking with Gemini for all the extra benefits it offers, despite the fact that Grok 3 has proven to be better in many cases during my testing. Your mileage may vary, though, since your use case may be different from mine. I wouldn’t dismiss Grok just because it’s not as talked about as its competitors, so try it out for yourself if you can to see if it suits your workflow better than rival chatbots like Gemini.