Perplexity AI Challenges OpenAI And Google With In-depth Research Tools

JAKARTA Perplexity AI is getting more and more serious in developing artificial intelligence-based research by launching a new tool called Deep Research. This technology is designed to provide expert-equivalent insights in a short time. Perplexity AI, founded by former OpenAI researcher Arvind Srinivas, is now a strong player in the AI industry with the support of big figures like Jeff Bezos and Nvidia.

This latest move directly challenges the expensive OpenAI research agent, offering a faster and more accessible freemium alternative. If OpenAI sells similar services for $200 per month, Perplexity provides free Deep Research with certain restrictions.

Paid users (Pros) only need to spend 20 US dollars (Rp. 320 thousand) per month to gain access to 500 deep searches per day. Currently, this service is available on the web and will be coming soon on iOS, Android, and Mac.

Deep Research Performance And Excellence

Technically, Deep Research is designed to emulate human approaches in understanding complex topics ranging from search, reading, to compiling reports complete with reference sources. These tools cover a wide range of areas, such as finance, marketing, health, to product research.

Even more interesting, Perplexity can produce reports in just three minutes, much faster than OpenAI which takes 5 to 30 minutes.

In testing "Humanity's Last Exam," the AI benchmark for expert equivalent analysis, Deep Research scored 21.1%. This figure exceeds other models such as Gemini Thinking (6.2%) and GPT-4o (3.3%), although still slightly below OpenAI which reached 26.6%.

Despite the technological advances, Perplexity AI faces legal challenges from major media regarding alleged use of unlicensed content. As a solution, the company has begun to forge partnerships with publishers such as Time and Fortune in revenue sharing schemes.

With AI's growing research market, Perplexity's strategy focused on speed, affordability and accessibility has the potential to make it an attractive choice for users who are reluctant to pay dearly for a class-enterprise AI solutions.