Perplexity AI has more than 1 million users in India, says co-founder Aravind Srinivas

Artificial intelligence (AI)-powered search startup Perplexity AI has more than a million users in India, co-founder Aravind Srinivas told Moneycontrol, reflecting the growing popularity of the service that has become the talk of the town in Silicon Valley in recent weeks.

"We want that number to expand to like 10-100X more because the Indian population is so big," Srinivas said in an interview. The answer engine, which aims to take on tech giants such as Google and Microsoft, currently has over 10 million active users across the world.

Srinivas said that the awareness of the Indian public to generative AI is quite good. "A lot of people are aware of ChatGPT and large language models...They are already using these technologies in creative ways," he said.

Watch the full interview

"So, I think we are going to see a lot more faster movement in the AI revolution in India compared to the previous web and mobile revolution, where we were a little behind the rest of the world," Srinivas said.

Founded in August 2022 by Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski, Perplexity AI launched its answer engine in December 2022. The company recently raised $73.6 million in Series B funding at a valuation of $520 million from marquee investors, a list that includes Nvidia, Amazon founder Jeff Bezos, Institutional Venture Partners (IVP), and Shopify CEO Tobi Lutke.

Also read: Google created a terrible internet; we can create a win-win solution: Perplexity AI’s Aravind Srinivas

Google's AI head Jeff Dean and former YouTube CEO Susan Wojcicki have also backed the company in a personal capacity. This investment brings the total funds raised by Perplexity AI to about $100 million to date.

Unlike traditional search engines, Perplexity AI offers a chatbot-like interface that allows users to ask any question and receive concise answers, backed up by a curated set of sources that are mentioned as citations. People can then ask follow-up questions or head over to the articles mentioned to dive deeper into that specific topic.

To be sure, Google introduced its generative AI search experience as an opt-in experiment to users in select countries including India last year. The feature provides quick summaries for user queries on top of search results.

The search giant also rebranded its AI chatbot Bard as Gemini along with launching a paid tier and mobile apps. The move comes as Google parent Alphabet positions Gemini as the main brand for all its existing and future AI efforts.

That said, what's particularly interesting about Perplexity AI’s answer engine is that people can narrow down their sources for their queries to specific ones such as academic journals, YouTube videos, Reddit posts or computational knowledge engine Wolfram Alpha.

Growing its India base

To accelerate the growth of Perplexity AI in the country, Srinivas said they need to further improve support for Indian languages and make the app  better for international markets. For instance, he mentioned that when users ask questions in their native language, the answer engine sometimes responds back in English language.

"We need to address these problems…Once we do that, I believe that we can certainly command a bigger user base in India. India is the next generation economy" he said.

In the future, Perplexity AI will also look at lowering its subscription price to a rate which will be better suited for the India market. Although it is unlikely to be very cheap, Srinivas said.

"India's GDP is growing so much that I believe the optimal pricing is something we will arrive in some time to come. I don't think people are going to be optimising for prices, like how they thought about Netflix or Spotify, because people are going to think that this tool makes me more productive at work, let me use it. Hence, there will be people who would be willing to pay up for the value this product adds to them" he said.

Perplexity AI currently charges $20 per month for a premium subscription with advanced capabilities, such as the ability to upload images and documents; the option to choose between AI models such as OpenAI's GPT-4, Anthropic's Claude 2.1, or the company's own experimental AI models; and $5 of API credits to use every month.

The subscription also allows people to access an interactive search companion Copilot (not to be confused with Microsoft's Copilot), which offers a more personalised search experience by taking user's preferences into account and fine-tuning its answers based on their needs.

Collaborating with Indian LLMs

Going forward, Perplexity AI also plans to collaborate with Indian startups such as the Lightspeed and Peak XV-backed Sarvam AI that are developing large language models (LLMs), which would better suit the country's diverse linguistic culture.

Also read: India's LLM race is heating up! Here's a look at who's building what

"I certainly hope that the models you train specifically for India ends up being a better model to use than the generic model like GPT or Claude or the models that we build. So it makes a lot of sense to just use that as a default experience for Indian apps" Srinivas said.

"I'm excited to be a customer of LLMs built by Sarvam or other Indian language model companies. When they're ready for go-to-market, we would love to collaborate with them and work on, because we are a big platform for them too. If we use their models, we have a user base that they can already bootstrap from, so we are excited to collaborate with them" he added.

Srinivas also said that venture capitalists (VCs) need to take bold bets on Indian founders who are trying to disrupt the existing incumbents.

Srinivas mentioned that there are two categories of companies that investors can back. One is people who are using existing APIs to build great product experiences for end users. "They don't need to worry too much about having your moats. Just get the user base in India and create the economy at scale for that user base. And then figure out how to optimize costs and build your own infrastructure or partner with other Indian infrastructure companies" he said.

The second set are the infrastructure builders, those that are building their own LLMs, their own hardware, inference chips, training chips, and data centers among others, which is a longer-term commitment.

"People should take bold bets on both of them, and the government should also be involved if possible. That's how you will see a lot more activity, like founders who will prefer to stay back (in India) and build companies for Indian users and the Indian ecosystem" he said.