Google’s AI Mode Lands in India—But It’s Not Ready to Replace Search Just Yet

For years, Google Search has been the place we start our questions-quick, reliable, and familiar. But that familiar white box is evolving. As AI reshapes how we search and discover, Google is now rolling out its experimental AI Mode in India via Search Labs, following its initial release in the U.S.

AI Mode promises a more intelligent, context-aware approach to search, allowing users to ask deeper questions using text, voice, or images. I spent some time with it, and while there's clear potential, the current experience feels like a work in progress.

Getting Started with AI Mode

Enabling AI Mode is simple enough:

• Open the Google app on your Android or iOS device

• Tap the Labs icon in the top-left corner

• Select AI Mode and tap "Turn on"

• Follow the setup prompts (you can also use AI Mode in your desktop)

The feature is part of Google's broader Search Labs experiments, and access is being rolled out gradually. There's no public timeline yet for a full launch.

AI Mode Feels Familiar-Because It's Gemini, Just Embedded

Once enabled, you'll notice a new tab appear on your search results page for AI Mode. It's not a radical UI overhaul, but a separate, cleaner interface where you can ask long-form or exploratory queries and receive synthesized responses.

Underneath, it's powered by a customized version of Gemini 2.5, the same model behind Google's standalone AI assistant. Unsurprisingly, then, AI Mode responses behave a lot like what you'd get from Gemini-structured, conversational, and context-aware. In fact, right now, it feels more like Gemini built into Search, rather than a completely new experience.

Product Queries Still Need Work

One of the first things I tried was a practical buying question: "What are the best phones under ₹30,000 with good cameras?"

AI Mode responded with a full list of suggestions, but the results were mixed. A few phones on the list were recent launches from this year, but several were clearly outdated, like the Pixel 6a from 2022 and the Galaxy S20 FE, which dates all the way back to 2020.

This wasn't just a matter of age. Many of the older phones included aren't easily available anymore, and they've long been surpassed by better options with newer chipsets, improved cameras, and longer software support. It was a confusing mix-some current, some irrelevant.

I followed up by clarifying, "Only show phones launched this year," expecting a more filtered result. But even then, AI Mode continued to include older phones alongside a few 2025 models.

That response made it clear: AI Mode currently lacks a reliable grasp of what's both current and competitive. It doesn't prioritize availability or relevance, and for product research, especially in categories that move quickly like smartphones, that's a serious limitation. At this stage, I wouldn't rely on it to guide a real-world purchase.

Where It Shines: Open-Ended Advice and Follow-Ups

The more promising side of AI Mode reveals itself when you move away from product suggestions and focus on open-ended or decision-based questions.

For example, I asked, "Should I buy a diesel car or a petrol car?" The response went beyond technical specifications. It offered a balanced breakdown of both fuel types, and importantly, factored in personal variables-how often I drive, daily travel distance, and long-term ownership costs.

I followed up by saying, "I drive rarely," and the AI responded that a petrol car would be more suitable, citing lower upfront costs, better fuel efficiency for occasional use, and how diesel vehicles make more sense for high-mileage users.

I also tested it with "Should I buy a prebuilt PC or build one myself?"-a common question among casual and hobbyist gamers. The response was solid. It laid out pros and cons for both routes, touching on budget, customization, ease of support, and long-term upgradability. It felt very similar to Gemini's style of answers: clear, neutral, and genuinely helpful.

This kind of adaptive reasoning and follow-up is where AI Mode begins to feel like more than just a smarter search box. When it works, it offers clarity that actually helps you move forward in your decision-making.

Multimodal Input-Still Not Fully Live

Google has advertised AI Mode as multimodal, meaning you should be able to input queries via text, voice, or images. The activation screen even shows a Lens icon in the search bar. But during my testing, that wasn't available.

I could type and speak, but I couldn't upload images or trigger Google Lens through AI Mode. This might be a staggered rollout, so I have yet to test the multimodal functionality.

A Quiet Shift in Search, Not a Reinvention (Yet)

It's worth keeping expectations grounded. AI Mode doesn't overhaul Google Search. It adds a layer-an optional mode for longer, more nuanced queries. It builds on the AI Overviews introduced last year but takes things further by synthesizing data, offering follow-ups, and (sometimes) adapting to context.

Google says early users are submitting queries two to five times longer than traditional searches. That aligns with my experience. Once you realize the system can handle detailed questions, you naturally start asking bigger ones.

But the system still fumbles with real-world complexity. It needs to get better at understanding what matters today, not just what exists in its training data.

Final Thoughts

AI Mode is an interesting addition to Search-but it's clearly not finished. It works best when offering guidance or structured comparisons. It's far less reliable when you need up-to-date, accurate information about fast-changing topics like tech products.

For now, think of it as Gemini inside Google Search, with some added conversational flow. It can be helpful, especially for open-ended questions, but it's not something I'd rely on fully just yet.

Still, it's a strong signal that the way we use Google is evolving. And Google knows it can't stand still.