Born in the 1990s, with a net worth of 130 billion in 3 years.

He rejected Meta's acquisition offer and ignored Apple's proposal. In just three years, he built a top - tier unicorn valued at 130 billion yuan.

Recently, in Silicon Valley across the ocean, a young man has become quite famous. He is full of negative energy and says that since he started his business, he has never had a good night's sleep. Because as soon as he closes his eyes, he feels that "those big companies" will steal his ideas. And indeed, in December 2022, a flagship product they launched was copied by Google, OpenAI, and Anthropic within a month. He was helpless and also found that big companies would do everything possible to prevent him from "speaking out".

Angry, he decided to actively participate in various entrepreneur forums, training camps, and round - table discussions. He constantly called on entrepreneurs to unite, turn fear into "motivation", and strive with all their might to break through the shackles of big companies.

However, just when you think this is a reflection from a "loser" or a "cruel story" from a grass - roots entrepreneur, you'll suddenly find that this young man is more like a monster. Despite rejecting Meta's acquisition offer, ignoring Apple's proposal, and publicly mocking Google, he managed to open up the market and built a top - tier unicorn valued at 130 billion yuan in just three years.

This is the story of Aravind Srinivas, the founder of the well - known AI unicorn Perplexity, and his two other post - 90s partners. Recently, according to media reports, they completed a new round of financing. The currently known information is that the overall scale is 100 million US dollars, and the valuation has reached 18 billion US dollars (nearly 130 billion yuan in RMB) - and it has only been two and a half years since they launched their first product.

A "Defector" from Google

In terms of the most practical "number of financings" and "financing scale", Perplexity is undoubtedly one of the top unicorns in Silicon Valley today. Since its establishment in 2022, they have completed a total of eight rounds of financing. The list of shareholders is very impressive. At the individual level, there are Jeff Bezos, Yann LeCun, Garry Tan, the current head of Y Combinator, and Nat Friedman, the head of the most aggressive AI fund NFDG. At the big - company level, there are NVIDIA and Databricks.

In terms of specific financial data, according to some disclosed information, Perplexity's ARR (Annual Recurring Revenue) is slightly less than 100 million US dollars. Therefore, after the completion of this round of financing, it means that Perplexity has obtained a valuation of 180 times its revenue. This figure far exceeds the industry benchmark of 21.8 to 40.6 times for artificial intelligence startups and 7.8 times for traditional SaaS companies.

Considering that their previous round of financing actually took place in May 2025, when the top Silicon Valley venture capital firm Accel invested 520 million US dollars in them at a valuation of 14 billion US dollars. A simple arithmetic calculation shows that Perplexity was actually able to complete two rounds of financing within two months at such a high multiple, and its valuation increased by 30%. Anyone who sees this will exclaim "amazing".

However, for a company like Perplexity, these numbers are rather underwhelming and ordinary. A better way to truly appreciate how great they are is to connect three pieces of news from June:

- First, at the beginning of June, Bloomberg reported that Samsung, which has been looking a bit lackluster in the era of artificial intelligence, decided to make a big move and heavily invest in Perplexity. After the investment, Perplexity will be directly integrated into Samsung's products. In exchange, Samsung needs to pay a large enough amount of money to become one of Perplexity's largest shareholders at present;

- In mid - June, Samsung's plan went awry because some media reported that Apple executives such as Adrian Perica, the head of Apple's mergers and acquisitions business, and Eddy Cue, the senior vice - president of Apple's Internet Software and Services business, were discussing the possibility of acquiring Perplexity. Forbes commented that this was Apple's "big bet", while Fortune magazine commented that this was the only way for Apple to save its "backward situation";

- At the end of June, Meta bought a 49% stake in AI unicorn Scale AI for more than 14 billion US dollars and poached Scale AI's core product team, including its founder Alexandr Wang, officially starting the "talent - grabbing" war in the artificial intelligence field. But some media reported that before finalizing the deal with Scale AI, Meta actually contacted Perplexity first, but Perplexity ultimately rejected the deal.

As the saying goes, if you want to see farther, you need to stand on the shoulders of giants. Nowadays, the giants are actually squatting down and bending over, actively showing their smooth shoulders, and competing fiercely, which clearly shows the quality of Perplexity.

So how did Perplexity achieve this? If summarized in one sentence: Perplexity is the Google of our era.

Aravind Srinivas, the founder of Perplexity, was born in Chennai, India, in 1994. As we all know, the IT industry in the hearts of Indians is like kimchi in the hearts of Koreans and pizza in the hearts of Italians. It is the crown that showcases national wisdom, and Chennai can be regarded as the pearl on the crown because it is the hometown of Sundar Pichai, the current head of Google. Growing up in such an environment, Srinivas regarded Sundar Pichai as his idol since childhood. Adhering to the spirit of "mastering mathematics, physics, and chemistry, one can travel fearlessly across the South Asian subcontinent", he embarked on the path of a problem - solver and successfully obtained a bachelor's and a master's degree in electrical engineering from the Indian Institute of Technology Madras, a well - known Indian institution.

In 2017, Srinivas decided to pursue his dream in Silicon Valley and went to the University of California, Berkeley, to continue his studies for a doctorate in computer science. This decision indeed gave him the opportunity to stand beside his idol. During his doctoral studies, he joined DeepMind as an intern researcher. After the internship, he was successfully converted to a full - time employee and participated in the development of machine - learning programs for computer vision at Google. Even today, the book that Srinivas reads most often is the corporate biography "In The Plex". He often says bluntly in media interviews:

"It was Larry and Sergey who made my entrepreneurial story possible."

However, what fascinated Srinivas was never the "name" Google, but the innovative spirit that shone at the beginning of its founding. By the time Srinivas, born in 1994, had the opportunity to join Google, the company had suffered from a severe "big - company disease", with numerous factions and a long - chain management structure. Especially in providing direct and context - relevant answers for complex information searches, Srinivas found that traditional search - engine products had shown considerable limitations, while the necessary optimization process was extremely slow and rigid.

Helplessly, Srinivas had to rethink his personal plan. After much hesitation, he reluctantly ended his career at Google and applied for a research position at OpenAI. At the same time, Srinivas also devoted more time to the study of machine learning, especially focusing on reinforcement learning and contrastive learning in computer vision. He quietly became a star at top academic forums such as ICLR, AAAI, and NeurIPS, and also quietly started the entrepreneurial story of Perplexity:

During academic exchanges, Srinivas met Denis Yarats, who had the same research direction and was working as a researcher at Facebook at that time. The two hit it off immediately and soon became close friends from online pen - pals who exchanged private messages. In early 2022, Denis Yarats invited his former colleague at Quora, Johnny Ho (referred to as Lao He for short), to join the group chat, and they began to discuss the possibility of starting a business together. In July 2022, the three post - 90s decided on the product direction and aimed to develop a search engine that fully integrated AI technology to make "information search" truly "people - centered".

(From left to right are Lao He, Srinivas, and Yarats, from social media)

In December 2022, the trio launched their first product, Bird SQL. This product was very rudimentary. Its overall framework was based on OpenAI's Codex tool, attempting to convert the daily language of Twitter users into SQL statements to better complete the Twitter search plugin. According to Yarats, Bird SQL was essentially a prototype. It was a quick way for the team to verify their product idea with limited funds, and Twitter happened to be one of the few databases on the market that offered a free API and had a large enough sample size.

The result was a great success, validating the correctness of their product idea and the public's dissatisfaction with traditional search engines. Just two months later, in January 2023, with enough experimental data, the Perplexity trio launched a more formal product, Perplexity Ask, which claimed to be the "world's first conversational search engine". While ChatGPT was becoming extremely popular, this product reached the milestone of 2 million active users within four months. After this product, the trio's entrepreneurial journey truly began.

The "Faith" War

By now, careful readers should have noticed something amiss: Perplexity should be a very romantic entrepreneurial story. It starts with a young boy's dream, continues with a young man's perseverance, and reaches its peak when a group of brave people challenge the product giants. It is full of the Silicon Valley spirit and fully demonstrates how the emergence of social networks can bring more possibilities to the world and every ordinary person. Even the name "Perplexity" itself is an Easter egg.

"Perplexity" is actually a term in the field of artificial intelligence development. Its original meaning is "perplexity", which is a core indicator for evaluating the text - prediction ability of language models and is used to measure the "surprise" degree of a model when it encounters new data. An example of its use is: the lower a model's "perplexity", the higher its accuracy.

So where does the founder Srinivas' anger come from? To sum it up in one sentence, although the Silicon Valley culture gave birth to Perplexity, Silicon Valley has never been kind to it.

Although the first product, Bird SQL, launched by Perplexity in December 2022 was rudimentary, it was also quite innovative. On the one hand, Bird SQL clearly demonstrated Perplexity's product vision. They hoped that the search engine could directly provide the "best answer" and become an "answer engine" instead of making users sift through a vast amount of relevant information on their own. On the other hand, Bird SQL successfully utilized Twitter's large - scale and real - time updated database, while most AI search engines before that were based on static training data.

Therefore, as soon as Bird SQL was launched, Perplexity quickly became one of the hottest startup teams in Silicon Valley. Even Jack Dorsey, the then - owner and founder of Twitter, couldn't help but promote it and personally retweeted Perplexity's official tweet, praising it as "This is great". The "world's first conversational search engine", Perplexity Ask, was also rapidly developed in this atmosphere and officially launched in January 2023.

However, in February 2023, the situation suddenly changed. Twitter officially announced that it would no longer provide free API services, and the Bird SQL product line was directly aborted. Fortunately, at this time, ChatGPT emerged, and conversational AI became extremely popular. Problems such as excessive hallucinations and unclear information sources in conversational AI broke out concentratedly. Perplexity didn't take long to re - position the focus of product development and dedicated itself to building the best - experience "question - answering information platform".

The Series A financing, in which Yann LeCun and Nat Friedman participated, was also completed at this time. At that time, Srinivas was very optimistic and said: "When people search for answers to questions online, traditional search engines present a large list of links that may be manipulated by advertisers and search - engine optimization (SEO). Users have to distinguish website information by themselves, and this information often has accuracy issues. The mission of Perplexity AI is to completely solve these problems."

But before Srinivas' words had even faded, trouble came again. In March 2023, Google Bard (now named Gemini) launched a real - time internet - connected query function highly similar to that of Perplexity Ask. ChatGPT followed suit in May 2023, and then Claude also started developing relevant functions. What was even more fatal was that at that time, Perplexity didn't have its own LLM (Large Language Model), and its products were basically built on the OpenAI's GPT - 3.5 model and Microsoft Bing.

This was a survival crisis that Perplexity had never experienced or even imagined. After much deliberation, the founding team thought that there was probably only one way to solve the crisis: if you want to fight, I'll fight.

In January 2024, Perplexity completed a milestone - style financing. First, this round of financing helped Perplexity attract new investors including Jeff Bezos, NVIDIA, IVP, NEA, Databricks Ventures, etc., almost rallying half of the Silicon Valley technology circle from celebrities to big companies. Second, the scale of this round of financing reached 73.6 million US dollars, making Perplexity's cumulative financing amount exceed 100 million US dollars. Srinivas said that they finally had enough chips to "expand the team". Last but not least, they finally had the ability to develop their own LLM.

However, just when the product was about to get on the right track, trouble struck again. In October 2024, several media institutions such as Dow Jones & Company (the publisher of The Wall Street Journal), The New York Post, Forbes, and Wired magazine under the News Corporation in the United States filed various lawsuits and accusations against Perplexity, accusing it of illegally scraping content and causing potential losses of immeasurable scale.

Interestingly, in the lawsuit statement, Robert Thomson, the CEO of the News Corporation, wrote such a passage: "We admire principled companies like OpenAI because they understand that integrity and creativity are crucial to realizing the potential of artificial intelligence." People searched the news based on this passage and found that OpenAI had signed "very attractive" revenue - sharing contracts with several media institutions. For example, in the revenue - sharing contract it signed with Axel Springer, the parent company of Business Insider, OpenAI paid a share of "tens of millions of euros".

Even more darkly humorous, in November 2024, during the US presidential election, Srinivas said that he came here to study in 2017, and now the company's valuation has exceeded 100 billion, but he has never been able to get a "green card".

In short, you can indeed tell the story of Perplexity's 130 - billion - yuan valuation in three years in a romantic way, believing that their sudden emergence proves the vitality of Silicon Valley and the eternal pursuit of disruptors by venture capital. But please never ignore Perplexity's bumpy growth experience, which almost completely exposes the difficulties that Silicon Valley entrepreneurs may encounter - here is not an open place; it follows the more primitive law of the jungle. There is resource monopoly, cost out - of - control, and big companies are increasingly capable of erasing all innovation advantages. Social responsibility increasingly overrides corporate survival, forcing entrepreneurs to face many out - of - syllabus propositions.

To some extent, Perplexity's experience also well explains why Meta can easily promote its "talent - poaching wave" and why Perplexity actively promoted a merger with TikTok in January this year. After all, for Perplexity's entrepreneurial journey, it may be difficult to wait for an open content ecosystem like TikTok again.

Perplexity may be waging a war of faith about the Silicon Valley spirit. So in a sharing session at Y Combinator in June this year, when Srinivas recalled his experiences over the years, he chose to tell all the Y Combinator students present in a very negative way: "If you believe that your company can generate hundreds of millions or even