Published Date : 04/06/2025
Perplexity AI, once a scrappy startup in the shadow of OpenAI and Google, is gaining traction in the generative artificial intelligence (GenAI) race. With a string of deals spanning Samsung, PayPal, SoftBank, and others, as well as reported talks with Apple, the company is aggressively expanding its footprint across both enterprise and consumer markets.
Its most recent partnerships include the following:
Perplexity AI is in advanced discussions with Samsung to integrate its AI capabilities into Samsung’s smartphone ecosystem. This collaboration may lead to Perplexity’s AI replacing Google’s Gemini assistant on Galaxy devices, with integrations planned for Samsung’s web browser and Bixby assistant.
Perplexity has partnered with Motorola to integrate its AI technology into Motorola smartphones, aiming to enhance user experience with AI-powered features.
Perplexity has added PayPal as a checkout option in its chatbot.
Perplexity has expanded its strategic partnership with SoftBank, who is an investor. SoftBank’s sales team will market Perplexity’s Enterprise Pro plan to corporate customers in Japan.
Perplexity has partnered with Wiley to offer users access to the publisher’s content within the Perplexity AI chatbot.
PYMNTS Intelligence data shows that companies are stepping up their investments in AI, with 90% of CFOs reporting seeing “very positive” ROI in December 2024, triple the number of respondents who said the same nine months earlier. They are using GenAI to enhance productivity rather than to replace human workers entirely.
Perplexity’s strategic alliances position the AI startup for growth despite not enjoying the household name recognition of Sam Altman, OpenAI’s CEO, nor the cachet of the Google and Microsoft brand names. When the White House under both Biden and Trump invited the who’s who in AI, Perplexity CEO Aravind Srinivas was not among them.
So why are large enterprises inking deals with Perplexity? “Perplexity has shown that being a model ‘polyglot’ can create more strategic advantages than owning the best single model,” said Dev Nag, founder and CEO of QueryPal, who had worked at Google and PayPal.
Unlike ChatGPT and other larger AI rivals, Perplexity’s chatbot doesn’t just use its own large language model (LLM) to generate answers. It can also switch to ChatGPT, Claude, or its own Sonar model. Depending on what the user needs, Perplexity brings up the right LLM for the task.
That means it’s user-focused. “While everyone fixates on who builds the smartest AI, Perplexity discovered that aggregating multiple LLMs and letting users hop between GPT-4, Claude 3.5, and their own Sonar model delivers better economics and resilience,” Nag explained to PYMNTS.
For example, when OpenAI raises API prices or has an outage, Perplexity routes the traffic to another LLM. “This approach lets a 150-person startup handle 400 million queries monthly without the crushing compute costs that would bankrupt most teams trying to run frontier models end-to-end,” Nag said.
That’s why enterprises like Perplexity. “The enterprise partnerships stem from this technical flexibility combined with something OpenAI initially missed: Citations and transparency actually matter more to institutional buyers than raw but untraceable intelligence,” Nag said.
Perplexity understands business needs. “Perplexity’s enterprise savvy comes from understanding that B2B adoption follows completely different rules than consumer viral growth,” Nag said. “While ChatGPT conquered hearts and minds through pure capability, enterprises needed the boring stuff first: SOC-2 compliance, data residency guarantees, and audit trails.”
Darren Kimura, president and CEO of AI Squared, told PYMNTS that Perplexity and OpenAI’s ChatGPT represent two very different approaches to the evolving AI landscape. “Perplexity positions itself as a real-time answer engine to provide concise, cited responses. This emphasis on provenance can appeal to users who are seeking verifiable information, such as researchers and executives,” Kimura said.
In contrast, ChatGPT serves as a general-purpose AI assistant, excelling in long-form reasoning, creative tasks, and brainstorming, with its strength in its retention and contextual understanding, Kimura said. Due to these different approaches, they appeal to different audiences.
Meanwhile, Perplexity is still dwarfed by OpenAI. ChatGPT saw 4.5 billion web visits in April 2025, compared to Perplexity’s 125.4 million, according to statistics from Semrush provided to PYMNTS. DeepSeek came in second, with 419 million visits, and Google’s Gemini third, at 133 million.
But Perplexity doesn’t need to be the biggest player to win. It’s positioning itself as a research-focused alternative to general-purpose assistants, with a business model built around transparency, multimodel agility, and actionability. “Perplexity’s bet is that fact-checking transparency and agentic capabilities can capture enough high-value queries to build a sustainable business, even without toppling the giants,” Nag said.
Q: What is Perplexity AI?
A: Perplexity AI is a startup in the generative artificial intelligence (GenAI) sector that has gained traction through strategic partnerships and a unique approach to business needs.
Q: What are some of Perplexity AI's recent partnerships?
A: Perplexity AI has partnered with Samsung, Motorola, PayPal, SoftBank, and Wiley to integrate its AI technology and enhance user experiences across various platforms.
Q: How does Perplexity AI differ from OpenAI's ChatGPT?
A: Perplexity AI positions itself as a real-time answer engine with a focus on concise, cited responses, while ChatGPT serves as a general-purpose AI assistant excelling in long-form reasoning and creative tasks.
Q: Why are enterprises choosing Perplexity AI?
A: Enterprises are choosing Perplexity AI because of its technical flexibility, ability to switch between multiple large language models (LLMs), and its focus on transparency and citations, which are crucial for institutional buyers.
Q: What is Perplexity AI's business model?
A: Perplexity AI's business model is built around transparency, multimodel agility, and actionability, focusing on delivering high-value queries with fact-checking and provenance.