ChatGPT vs Google Gemini vs Grok - AI Comparison After The Hype...

I fused them all. I really did. I tested all LMS, both big and small. My top three right now are Chad GBT, Gemini, and Rock. Each is great for a reason, whether it's an exciting package, promising features, or a huge potential. But which one should you pick for daily use and your business? In this video, I'll compare all of them, show you what they can and can't do, and how to use them to their fullest.

This will be an unusual comparison. I won't just go over facts or compare the responses to the same prompt. Instead, I'll compare my experience with them, the little things I noticed, and I'll assign points, you know, rating. In each category, I'll have things to add points for consistency in images, quality, research, etc. For example, for the same feature, one tool can get nine points because it's almost perfect, another one five points, and another tool zero. Then I'll sum them up and announce the winner. Thank you Hrefs for sponsoring this

this video. The GBT isn't officially out yet. That's true, but all the leaks point to the same idea behind it. Taking all those models we have right now and combining them into one. No separate model for reasoning or for text generation or for working with large context inputs. One model to roll them all. In this idea, Grock is the closest one right now. You only have one model and all the toggles for research or reasoning only turn on the features themselves without changing the model. Gemini does the same thing but needs to switch a model for research. And Chad GBT right now, you know how it is. So, right off the bat, I'll give Grock two points, bonus points, and Gemini one point for outpacing Chad GBT. The number one skill of all LLMs is

writing. They all do it really well, but only if you start looking at small things, you start noticing the rift between them and different approaches they chose. Take style for example. All

three can slip into any voice you want. Shakespeare, newspaper reporter, or a hyper 5-year-old. The first reply is usually perfect. The trouble starts when you drop in a long chunk of text or shout change style halfway. That is when they wobble. Chad GBT is best here. grabs the vibe and holds on tight even after several edits. Its weak spot is memory. Tell it to shorten a passage and keeps trimming every time until you end up with one lonely sentence. Gemini also shranks text, but not as badly. It sometimes misses the vibe or refuses certain changes. Rock is the oddball. It's not fancy with style and sometimes ignores orders. Yet, it stays steady and predictable. So, Chad GBT gets eight points here. Gemini gets six points and Grog gets four points. Facts matter too. Every model can search the web, but only Gemini does it by default. With Grock or Chad GBT, you must flip the used internet switch. Forget they rely on all training data, which is not perfect or fully checked. In some of our use cases, correctness is especially important, like optimizing videos for search, writing descriptions, and so on. Neither of the three tools can do that correctly. So don't even try them for optimizing your website or any online presence with them. When creating our website aim masteraster. me we've used hrefs for optimizing the website for search and we still have a long way to go. I just add my website pick competitors keywords and see the whole stats. The tool brings all my keyword content and brand checks into one place. It runs on big data, so every report loads fast and feels deep, and I can install an add-on onto my website for even better data crawling and even deeper insights. That depth lets me spot rank and wins and losses without guessing. But a website is just a website. And for ranking high, we need to know how search engines and AI see our website. For this, I use the brand radar feature. It shows how chat bots and other AI answers talk about my brand. It tells me when my brand appears, how it stacks up to rivals, and where new chances open. If the tool sees that I'm missing a theme the market cares about, I learn early and can publish or promote content before traffic slips away and uh I can write copy or optimize the existing one with a built-in AI content helper. It turns that data into stronger writing. It spots the real intent behind any keyword, rates drafts against the best pages in the topic, and points out every gap. When a sentence feels off, the ask AI prompt rewrites, shortens, or expands it on the spot. It can even suggest fresh metatitles and descriptions, so my work is ready for readers and search engines without extra steps. Together, these parts create one steady flow. HRES finds the topics. AI content helper shapes the story and brand radar checks how that story travels through both classic search and new AI results. The suite keeps my work focused, my ed is light and my brand easy to discover. Check out hrefs by clicking the link in the description. With search turned on, all three are about equally accurate. Their tone though is different. Chad GBT raps facts and bubble wrap very safe and neutral. Gemini sticks to numbers and citations which can feel stiff. Grock sits in the middle. Elon calls it truthseeking AI and it tries to live up to that. Still, always check the sources. Grock gets six points. Gemini gets five and ChBT gets four. That feels right. Role playing is only a small

slice of what these bots do, but it matters. When a model pretends to be a pirate or a math tutor, it narrows the info it pulls and totally shifts the tone, the words, the whole vibe. Most of the time, you don't even have to spell it out. The bot guesses what you want. Or so it claims. Chad GBT is best sustained character about 99. 9% of the time. It keeps the act going. You can knock it off course if you really push. But if you just keep asking for more of the same style, behaves. Just don't pile on a bunch of new formatting rules that scrambles his brain fast. And if one answer lands way off, open a fresh chat and move on. The first reply sets the mood. And if it's wrong, there is no risk in it. Every model has that flaw. So, for role play, I'd give Chad GBT eight points, Grock 7, and Gemini

five. Image generation is another feature you're going to use often, and each model can do it, but none of them is perfect, and you really need to know the weaknesses and strengths of them if you want good images. All three tools use the same basic setup, a prompt box

and a magic go button. But how you deal with the finished picture is different. None of them lets you pick the exact aspect ratio or resolution by hand, which feels like ordering fries but not choosing the size. Even the latest Imagine 4 update is like that, a bit of a letdown. Chad GBT and Gemini keep it simple. Type a prompt, get a picture, done. The Imagen 4 engine in Gemini is cool and sometimes can produce images as good as Chad GBT, but it's all super prompt dependent. Only Chad GBT and Grock let you jump back in and tweak things. With Gemini, what you see is what you get. Even the latest update doesn't add any new features here. Grock keeps edits easy. Pick a style, swap the background, or change the main subject with drop- down menus, then add a quick note if you need more changes. Chad GBT can do fancier edits. You can lasso part of the picture and tell it exactly what to fix. But the interface is clunky, like typing a password with a TV remote. So for ease of use, I'd give Chad GBT a solid nine, Grock 8, and Gemini an eight in advance for that new promising

update. When we talk about picture styles, one winner jumps out. Chad GBT. Its image engine is the newest of the three, so it can copy way more looks. We even dropped a haul video packed with fun styles you can try in Chad GBT. So check that later if you like. Chad GBT also remembers the style you pick better than the others. You can make a picture in say a pixel art, then ask for the same picture in watercolor and every tree, cloud, and coffee cup stays right where it was. Only the paint job changes. Grock can do this trick too, but the scene shifts more and sometimes ends up looking like a brand new picture. Gemini just starts over every time, which feels a bit random. Google really needs to tweak that part. So on my scoreboard, Chad GBT gets nine points, Grog takes seven, and Gemini style work lands a four. The hardest

test for any image bot is writing clear words inside the picture. I won't sugarcoat it. Chad GBT stops again. Gemini and Gro still stumble on basic words or mix up letters while Chad GBT can draw whole comic pages with speech bubbles and the same characters from panel to panel. The story reads fine, not like scrambled code. I've even had it spit out restaurant menus in two languages. And the spelling, the fonts, and spacing all look sharp. Just one tip. If you need a paragraph or two, paste your own text and tell a chat GBT where to put it because making it invent long text can still glitch. Even so, that's an easy 10. Gemini and Grock each get three points for

effort. Scoring brain power is tricky because the setups differ. Grock and Gemini each run one main reason and model. Chad GBT lets you pick from 03, 04 mini, 04 mini high or 01 pro. And they each behave a bit differently. To keep this fair, let's stick with 03. With 03 you get the full Chad GBT tool belt. can hop online for fresh info, dig deep in research mode, open a bunch of file types, and crunch numbers fast. Gemini thinks things through pretty well, too. Reasoning is always on, and you can switch it off. It handles everyday stuff fine, but I wouldn't hand it a big data set without double-checking the numbers. Grock can reason yet without that extra toolbox feels a step behind when the task gets heavy. That reasoning is just one of cool features. But the true colors of these LM show only if you push them to their limits like we did for all members. We have created special AI agents to help you learn AI better. Like here's an agent that helps you write better prompts. You give it a task and it gives you a prompt. Or here's an agent for finding AI tools. But my favorite one is the news agent. Gives you neat summary with all the details. Super easy to stay on top of things. And we also have a virtual version of me for you to talk to. All that and more is available for our members. So hit the link below and become a part of the family. With Grock, you actually have to hit the little thank button before it digs in. Leave that off and it just skims the surface. Even when think is on, Grock handles everyday stuff fine, but it runs out of steam on deeper puzzles, on big data sets, and on tricky search results. Chad, GBT, and Gemini are stronger here. Yet, all three get better if you give them a few simple nudges. Spell out the question in plain words. Tell the bot to look at the problem from two or three angles and ask it to quiz itself before it answers. Turn on each platform's deep research mode, and things climb another notch. But I'll compare those modes later. For now, my scoreboard lands at Chad GBTA, Gemini 6, and Grog 5. Feels fair to me. These bots do more than type, but each one branches out in its own way. Chad can talk and listen, stare at the photos you drop in and draw brand new pictures. Plus, it happily hands you finished files, images, documents, spreadsheets, you name it. It still can't watch videos, though. Gemini can. It sees short video clips as well as photos. And it can talk and listen, too. But it's stingy about giving files back to you. Grog keeps it simpler. It writes, rewrites, looks at images, and opens data files. No video and no file downloads. Because Gemini does see video, I give it eight. Chad GBT earns seven for his all-around file skills. And Grock scores six with room to

grow. Talking about the hidden weapon features is tricky because each bot keeps a different bag of tricks. Instead of crowning one winner, I'm just writing how helpful each bonus feels in real life. First up is Gemini with its audio overview button. You drag in a dock or a slide deck, press that one little button and boom, it turns the text into a snacksiz podcast. Two AI voices chat back and forth, hitting the big point, so you can pop in your earbuds and listen instead of squinting at the screen. The same thing works on your phone, so it feels like having your own pocket tutor or mini radio show. It's handy. It saves time. And for that, Gemini Pockets 7 points. Grock's extra features are slimmer. There's an autoplay button off to the side, and there are two search depths called deep and deeper. That metal depth lands right between a quick Google and full research dive. It spits out answers fast, chews through the length of polls, and I actually lean on it whenever I need a quick web scrape. It's basic, sure, but useful enough to score five points. Chad GBT shows up with two headliners, schedule tasks and operator. Schedule tasks are that simple. You tell the bot to do something at a set time. Maybe it grabs the morning news and feeds you a summary or pings you when it's time to clock in or clock out. The idea is cool, but right now the use cases are still thin, so that bit earns four points. Operator is the bigger play. Think of it as a little helper that clicks around the web for you. It opens sites, press buttons, fill out forms, and line up prices on whatever gadget you're shopping for. You get to watch its moves in real time. And the only step it won't take is pulling out your credit card for trip planning, deal hunting, or paperwork prep. It's a lifesaver as long as you're ready to jump in if it drifts off course. That usefulness gives operator and by extension chat GBT a solid seven

points. Some tools are shared across the three bots and the big one is canvas. Think of canvas as the moment the chat window puts on a costume and turns into tiny writing studio. But each model runs the show in its own way. In every case the screen splits in half. Yet the buttons you get are different. In Chad GBT, you can just swipe over a sentence, click once, and tell the AI to tweak it. Shorter, longer, simpler, maybe sprinkle in emoji. You can do that for a single line or the whole draft. Feels natural, like nudging a friend's Google Doc in real time. Gemini copies that idea almost beat forbe. Same split U, same row of buttons. The one bonus needs send to Google Docs link so your finished piece flies straight into Drive. Henny, sure, but I don't mind a quick copy paste from Chad GBT, especially since Chad GBT still writes the cleaner first draft. One catch, in Gemini, you have to flip canvas on by hand every time while Chad GBT sometimes turns it on by surprise. Pick your poison. Grock hides its canvas behind a secret. You have to type use canvas mode because the bot never guesses when you need it. Once it finally appears, the editing tools are super thin, bald, itallic, shuffle the heading. That's about it. You can grab a paragraph and say, "Rewrite this, make it funnier, or stretch or shrink the length. " It's surface level paint. No deep remodel. Whichever bot you pick, a couple of tricks help in Chad GBT or Gemini. Always highlight only the words you want fixed. The model stays locked on that spot and leaves the rest of your page alone. and play with the tone and length sliders before you press done saves you an extra back and forth. So stack it up and the pattern is obvious. Chad GBT gives the richest workspace even more so in its second canvas mode for code. Gemini lands comfortably in the middle with that tidy split view and the docs export. Grock keeps life minimal for now. Scoreboard Chad GBT clear 10, Gemini 8, Grock and honest four. Feel free to fight me in the

comments. We already hit files and multimodality, but I can skip data crunching. Thanks, sales reports, giant tables with hundreds of rows, the fun stuff. Every bot can open a spreadsheet and run numbers. Yet, the real gap is how easy that feels, how well the bot remembers what you just did, and how smart it is at reading the sheet. If you want a standalone tool, Chad GBT is still my top pick. Fire it up, toss in a CSV, and it starts charting without a fuss. Need something glued straight into your daily workflow? That's where Gemini shines because it lives inside Docs and Sheets. You don't upload anything? You tweak cells and launch the analysis right inside the file. The AI talks to your data in real time, which is slick for teams already married to Google Workspace. Personally, I lean Chad GBT for heavy lifting. It's not as deeply wired into outside apps, but the charts look clean. They are interactive, and it fixes badly formatted data on the fly. Want a tweak? Just ask. No rerunning the whole analysis. Grock, yes, it handles files and crunch numbers, but you need the pay tier, and that isn't cheap for most folks. There is no reason to drag rock into spreadsheet duty. So scores rock three, Gemini 8, Chad

GBT8. Deep research looks the same at first glance in Chad GBT, Gemini, and Grock, but each one acts a little weird when you hit the switch. In Chad GBT, the feature really does hide under a tiny button, and the moment I drop in a prompt, it fires back with follow-up questions. It keeps asking no matter how much detail I took into that first prompt. The digging can take anywhere from a couple of minutes to half an hour depending on how big the job is. Every now and then it glitches and shows nothing at all, which hurts if it happens after you've already run two or three long searches in the same chat. Deep search works on any tier, even the free one, but the free pass things a bit simpler and the answers feel a notch lower in quality. If you plan to lean on the feature a lot, spring for a paid plan. With Gemini, you have to swap over to different model first. Once you're in, the flow looks exactly like Chad GBT. Clarifying questions, then await. The output is just as wide and neatly structured. The cool bonus is memory. Gemini's buffer is a couple of times bigger than Chad GBT. So, the odds of hallucinating inside a long thread are much smaller. I can even run two separate deep dives in one chat and cross-ch checkck the results without losing the plot. Grock, as I said before, offers two research gears, deep nails, quick lookups, while deeper mirrors the same full dig you get in Chad GBT and Gemini. Same weight time, same depth, but Grog never throws clarifying questions. Either it's reading my mind or it's simply sure of itself. And honestly, I like that. Those extra questions wear me down, especially when the bot could sort out half of them alone. For that reason, I'd hand Grock an eight here, while Gemini and Chad GBT both land at seven. I know I already talked about

special features, but I just can't ignore the smaller custom models. In Chad GBT, they are called custom GBTs, and in Gemini, gems. Gro simply doesn't have features like that, so it gets a flat zero point. Google's gems are quite cool. There are a couple made by Gemini, like a gem for coding, helping with your career, or coming up with ideas, and it's really easy to create one yourself. Simple prompt and a knowledge base are enough. The controls are fairly simple, but reliability is pretty decent. That's worth seven points in my book. Though, gems aren't even remotely as powerful as custom GBTs because those can do much more. For starters, they are just easier to create. You can do it through a chat or through a special configuration window where you not only give it instructions and knowledge, but also enable or disable some of its features, but these are just the basics. You can make GBTs that will access third-p partyy resources and platforms. People have been going crazy with these this past year. There are special GBTs for Canva, Wallframe, Scyace, and many more. You can find a custom GBT for any task specifically trained on that sort of thing. I do wish these custom GBTs were even more versatile in terms of their integration into third party services and platforms, but even the stuff we have now already deserves a

nine. Let's draw a line, right? Chad GBT 108 points, Gemini 83 points, and Grock 68 points. I'm honestly surprised because in real world use, all models feel very similar. But that's exactly what I said earlier. Unless you pay attention to the details, you won't notice a difference, and details matter. For each of these three models, we already have detailed guides into all their features. Be sure to check them out. And don't forget about our 101 crash course of Generative AI at Geek Academy. And if you want to sponsor one of our videos, send us an email collabinerdia. com. This is the only real email. Thanks for watching. I'll see you in the next video.