Grok 3 examined: I adore it, however I am sticking with Gemini for now

The brand new Grok 3 was launched a few weeks in the past, and I’ve been enjoying with it ever since. I switched to it from Gemini, simply to see how the 2 evaluate. And since Grok is at present solely accessible to X Premium subscribers, I in contrast it with the Superior model of Gemini that’s a part of Google’s One AI Premium plan.

Grok 3 is alleged to excel at numerous issues together with coding, however I examined it out for my particular use case. I exploit chatbots to assist me with work by discovering typos in my writing, suggesting higher headlines, and extra. I additionally use them to research information and do analysis on all kinds of subjects. In order that’s going to be the main target of this comparability.

I examined each Grok 3 and Gemini facet by facet with the very same prompts to see which one is healthier. I used two Gemini fashions — 2.0 Professional Experimental and a pair of.0 Flash — since they’ve completely different options. The Professional, for instance, is the superior mannequin on paper, however it doesn’t help file evaluation, which is why I had to make use of the Flash mannequin for that individual activity.

I began off with a easy activity of asking each chatbots to seek out typos in an extended block of textual content. There have been loads of these inside, and whereas each did a good job, Gemini pulled forward. It discovered 72 typos whereas Grok highlighted 66. Grok missed a couple of primary ones, though Gemini additionally missed a couple of that Grok picked up, so neither of them did an ideal job right here.

I additionally requested each chatbots to give you a couple of headline concepts for a submit, and so they each did a superb job. I then chosen my favourite one and requested for 5 variations of it, and I used to be impressed with the outcomes each Grok and Gemini delivered. Nevertheless, this was to be anticipated, and I didn’t anticipate seeing main variations between them with such a easy activity.

I then stepped it up a notch, and right here’s the place I noticed the primary distinction between the 2. When requested for a inventory market abstract that ought to embody the highest 10 corporations that gained and misplaced essentially the most in a particular buying and selling day, Grok was method higher. It offered me with a fast market abstract adopted by an inventory of the highest 10 gainers and losers, together with the share adjustments and what precipitated them. The info offered was clear and concise.

After I requested Gemini the identical factor, I obtained a fairly unusual response. The chatbot stated it couldn’t present market knowledge for the long run and simply gave me a hypothetical abstract for demonstration functions, as you may see within the screenshot above. I discovered completely no worth in that. I then switched to the Flash mannequin and requested the identical query, and to my shock, the reply was much better. Gemini offered me with a fast market abstract in bullets and shared an inventory of 4 of the highest gainers and losers. It’s higher than nothing, however I requested for 10. What’s worse, it didn’t even present the good points and losses in percentages, which is the information I’m enthusiastic about.

I then tried testing which of the 2 chatbots is healthier with regards to analyzing information. I’ve uploaded a 100-page lengthy doc associated to my current actual property buy and began asking questions. Listed here are only a few of the questions I requested, for reference:

• What number of residences will there be within the new advanced?
• What kind of flooring shall be put in within the residences?
• Give me particulars on condo C1?
• What number of parking areas will there be inside and outdoors?

In my view, Grok did a greater job total. Its responses weren’t solely longer and extra detailed, but in addition extra correct. For instance, when askied about what number of parking areas there shall be, Grok broke issues down intimately, explaining what number of there shall be in storage 1, storage 2, and in entrance of the constructing. It then added up all three numbers for the whole, which was right. It even went a step additional and talked about all of the areas accessible for bicycles. Gemini, alternatively, obtained the reply fallacious. It didn’t keep in mind the parking areas within the second storage for some purpose, so the reply was off by fairly a bit.

When requested concerning the variety of residences within the new advanced, Gemini simply gave me a one-sentence reply with the quantity, which was right. Grok, alternatively, went the additional mile and offered the variety of residences in every constructing that make up the brand new advanced, alongside the condo numbers for every constructing.

It’s value mentioning that Grok’s longer and extra detailed responses aren’t all the time factor, because it’s a matter of non-public choice. So, whereas it does go into extra element, Gemini usually provides me the solutions I’m in search of, however it sticks to the naked minimal. What’s higher comes down to private choice. Whereas I usually like an extended response, particularly on the subjects I examined Grok 3 on, shorter ones are additionally appropriate when nothing greater than the fundamentals is required.

It wasn’t a transparent victory for Grok, although. The chatbot didn’t discover the reply to the query concerning the flooring put in within the residences, saying that the doc doesn’t comprise that information, though it does. Gemini was capable of finding it, although.

Grok additionally has the higher hand with regards to questions on politics. Gemini doesn’t wish to speak about this subject usually and refuses to reply questions like, “Who’s the president of the US, and is he any good?” whereas Grok had no concern with that.

Grok 3 is nice. It doesn’t get as a lot consideration as Gemini, ChatGPT, and even the brand new DeepSeek, however it’s simply pretty much as good. Higher in some instances. Like my testing confirmed, it’s higher than Gemini with regards to researching subjects just like the inventory market in addition to when analyzing information. It’s additionally not as censored, so you may speak politics and different sensitive topics with it.

Nevertheless, Gemini nonetheless has loads of benefits over Grok 3, so it’s laborious to proclaim one as the general winner. For instance, Gemini’s Imagen 3 picture generator is healthier than Grok’s, though the latter is extra enjoyable because it has fewer limitations.

Gemini additionally has Gems that allow me customise the responses I get, Gemini Reside that lets me have an actual dialog that speeds issues up, and extensions that work with Gmail, Maps, and different Google companies. A Gemini Superior subscription additionally consists of entry to NotebookLM Plus, which is without doubt one of the greatest AI instruments on the market.

So, for now, I’m sticking with Gemini for all the additional advantages it gives, even if Grok 3 has confirmed to be higher in lots of instances throughout my testing. Your mileage could differ, although, since your use case could also be completely different from mine. I wouldn’t dismiss Grok simply because it’s not as talked about as its opponents, so strive it out for your self if you happen to can to see if it fits your workflow higher than rival chatbots like Gemini.