I gave Google’s AI Mode a shot, but I’m still sticking with Perplexity

We’ve already seen AI creeping into Google Search, whether you liked it or not. AI Overviews started off as a frustrating product that marred the original Search experience with slow responses, inaccuracies, and occasional walls of unhelpful text. But it slowly got better, and for quick queries where you didn’t want to scroll through forums or click on five different links, it actually became useful. That’s what made me hopeful when, at I/O, Google announced a full-fledged AI Mode for everyday use.

Now, I’ve relied on Perplexity for my AI-powered web searches for a while. I stopped “googling” a lot of things long ago. So, if the same kind of AI capabilities could be baked into Google Search, why would I use another app, right?

Turns out, it’s not that simple. Once I actually started comparing the two, the results were far more complicated.

To give both tools a fair shot, I asked them the same set of questions to see how well they performed. With Google’s massive understanding of the web — and the fact that most of the internet is basically designed to work around Search — I assumed AI Mode would wipe the floor with Perplexity.

But here’s the thing: it’s not about how much data you have, it’s about what you do with it.

Logical lapses loom

We talk about smartphones all day over here, so of course that’s where I started. I imagined an average user who wants to upgrade their phone without doing intense research — reading long reviews, watching YouTube videos, and double-checking Reddit posts for credibility. So, I asked AI Mode to help.

The prompt was simply to suggest the best phones from Samsung and Apple under $1,000.

Google’s AI Mode confidently came back with the Galaxy A56 and the iPhone 16e — solid phones, sure, but they’re way below the budget. The rest of its suggestions weren’t any better. It thinks the iPhone 14, a refurbished iPhone 13, Galaxy A35, and the Galaxy S23 Ultra are good phones for that kind of budget. It was basically a mess with a confusing mix of old and new, premium and budget handsets.

Anyone relying on this to do actual market research would walk away misinformed.

I tried to guide it further. I insisted on the latest models and asked for a side-by-side comparison. I did get the comparison, but the suggestions remained stuck in budget territory. It just wouldn’t budge.

Perplexity, on the other hand, gave me a table in its first response — clearly laid out, with each phone’s standout features and who is the phone for. More importantly, it gave me what I actually expected for under $1,000: the Galaxy S25 Plus and iPhone 16/16 Pro. You know, the actual current-gen stuff.

Also, Google’s AI Mode felt like it was pushing for a sale. The entire results page was littered with shopping links like an overeager salesperson — that’s when I was still trying to decide what I even wanted.

I love that these AI tools can save me from opening 20 tabs just to get one clear answer or create a decent travel plan. I asked both AI Mode and Perplexity to help me plan a budget seven-day trip to Vietnam, and AI Mode was again a mixed bag.

Google’s AI Mode was definitely fast. It gave me a full day-by-day plan in the first result itself. The results from both AI Mode and Perplexity curiously felt too similar, as if they were learning from the same source material. But the presentation is where the gap was actually glaring.

AI Mode gave me a long, exhausting block of text broken up only by bullet points with their own blocks of text. So many words force you to skim the information that is already a summary of even more text, defeating the very purpose of intelligence, artificial or otherwise.

Perplexity needed one follow-up for a detailed plan, but the answer was so much better — clean tables, well-spread-out text, and a table with itemized and total estimated expenses. I mostly use Perplexity for work on my desktop, and the big-screen experience was even better. Each suggested hotel had a photo for reference, a map view of its location, and some quick pros and cons of the place (pictured below). That kind of little touch goes a long way. Meanwhile, the world’s biggest search engine struggled to even include images in its AI results despite explicitly asking for them.

AI Mode = Awkward Implementation Mode

I’ve already mentioned how text-heavy AI Mode feels, especially on a phone screen. The layout needs more love from Google’s designers because right now, it looks dense and, honestly, even a bit off-putting.

Beyond the visuals, the actual workflow is a tad clunky too. You have to go to the AI Mode section from the Google Search page or app, and only then can you start typing your query. You can’t start talking to AI Mode straightaway like Perplexity lets you. Add to that, the presence of AI Overviews and AI Mode on the Search results page, both trying to serve the same purpose: adding to the clutter.

And don’t get me started on chat history’s location — it’s again a multi-step process because you have to enter AI Mode first. AI Mode doesn’t give you an easy way to revisit past threads, which is kind of the whole point of AI search tools that work around follow-ups and context, which would get lost if you can’t continue in the same thread.

There’s a clear difference between a product built for AI from the ground up, and one where AI was just an add-on.

Perplexity feels more intentional and logically sound. The results are better laid out, the interactions are more fruitful, and it doesn’t seem to have a hidden motive, like getting you to click the “Buy Now” button. It’s not perfect, but it feels mature, like an old wine.

Google’s AI Mode is a little rough around the edges, but given it’s still a new feature, I can already see where it could reach with some polish. Knowing Google and the pace at which it’s iterating with its Gemini models and cross-app AI integrations, I wouldn’t be surprised if I came back in a few months and switched over completely.