Perplexity AI has introduced Deep Research, a new feature aimed at providing advanced AI-driven responses with real citations.
This tool positions Perplexity among established players like Google and OpenAI, both of which have recently launched similar research-focused AI tools.
Deep Research is designed for expert-level inquiries in areas such as finance, marketing, and product analysis, catering to users who require more than standard AI chatbot responses.
Unlike OpenAI’s Deep Research, which requires a $200-per-month subscription, Perplexity offers limited free access, with Pro subscribers receiving up to 500 queries per day.
Research Capabilities and Performance
Perplexity’s Deep Research tool operates iteratively, refining its approach by searching, analyzing documents, and adapting its methodology—an approach designed to mimic human research processes.
Users can access this feature via a drop-down menu and export results as a PDF or a shareable Perplexity Page.
Benchmark testing shows that Perplexity’s Deep Research scored 21.1% on Humanity’s Last Exam, an AI evaluation for expert-level academic tasks.
This score places it above Gemini Thinking (6.2%), Grok-2 (3.8%), and OpenAI’s GPT-4o (3.3%), though OpenAI’s own Deep Research tool leads with 26.6%.
Speed is another key factor where Perplexity distinguishes itself. Most queries are processed in under three minutes, compared to OpenAI’s Deep Research, which takes 5 to 30 minutes for similar tasks.
Comparison of AI Research Tools
Perplexity outlines three key differences among AI research tools:
- Perplexity AI: Prioritizes speed and accessibility for general researchers.
- OpenAI: Focuses on deep analytical capabilities, suitable for enterprise applications.
- Google: Integrates AI research with its broader productivity ecosystem.
While AI-driven research tools continue to evolve, some challenges remain.
Analysts, including those at The Economist, have raised concerns about potential over-reliance on AI-generated research, which may limit human creativity and favor easily accessible sources over deeper analysis.
Investment and Expansion
Founded in 2022 by former OpenAI researcher Aravind Srinivas, Perplexity AI has quickly gained industry traction.
Major investors, including Jeff Bezos and Nvidia, have backed the company, with $500 million in funding raised as of December 2024, valuing Perplexity at $9 billion.
The company has also explored strategic partnerships, such as a potential collaboration with TikTok U.S., which could expand its reach and influence within the AI research sector.
Freemium Model and Platform Expansion
Perplexity’s Deep Research feature follows a freemium model, allowing up to five free queries per day, while Pro users can conduct up to 500 queries daily.
Initially available on the web, Perplexity plans to launch the feature on iOS, Android, and Mac, broadening its accessibility.
This approach makes Deep Research more accessible to a wide range of users, from casual researchers to professionals who require frequent AI-assisted analysis.
Legal Challenges and Content Agreements
Despite its rapid growth, Perplexity AI has faced legal challenges from major media organizations. In October 2024, subsidiaries of News Corp, including The Wall Street Journal and The New York Post, filed a lawsuit alleging unauthorized use of copyrighted material.
The complaint claims that Perplexity’s AI-powered search reproduces large portions of journalistic content without appropriate permissions.
In response, Perplexity has initiated revenue-sharing agreements with publishers such as Time and Fortune, aiming to compensate media companies when their content is cited in AI-generated responses.
Perplexity’s AI Ecosystem
Perplexity is expanding beyond Deep Research, developing a broader AI-powered information retrieval ecosystem. Its free-tier model incorporates a custom GPT-3.5-based AI with browsing capabilities, enabling AI-enhanced searches with inline citations and summarized results.
For Pro users, Perplexity offers:
- API access
- Internal file searching
- Advanced AI models, including GPT-4 and Claude 3.5
These features indicate that Perplexity is evolving from a chatbot or search engine into a full-scale AI research assistant.
With its fast processing, accessibility, and citation-based approach, Perplexity’s Deep Research could emerge as a strong competitor in AI-assisted professional research.
As AI-driven analysis becomes more prevalent, the competition between AI companies in this space is expected to intensify.
