Perplexity joins the research assistant trend with the creatively named 'Perplexity Deep Research'
Perplexity has launched Deep Research, a free AI-powered research tool that can analyze hundreds of sources in minutes to create comprehensive reports across various domains, promising to save users significant research time.
Perplexity is the latest major AI startup to reveal an assistant capable of performing in-depth research from sources available on the internet and presenting its findings in a neatly organized report that can be exported into document formats. According to the startup, Perplexity Deep Research mimics how humans research a new topic, iteratively searching for and reading documents, and adjusting its strategy as it learns more. The main difference is that in 2 to 4 minutes Perplexity Deep Research can accomplish tasks that would usually take hours for a human.
Through a host of examples comparing Perplexity's default search with Deep Research, the company demonstrates the products research capabilities across topics as varied as finance, marketing, technology, current affairs and biography. Perplexity claims that its Deep Research product works as well at creating work artifacts from specialized marketing, finance, and technology research as it does as a personal consultant for health, product research and travel planning.
Moreover, the startup provides some benchmarking results, as this has become a more or less standard method to evaluate LLMs and LLM-based products. On Humanity's Last Exam, the newest comprehensive benchmark testing for expert-level knowledge with a closed-end question format, Perplexity Deep Search achieves 21.1% accuracy, second only to OpenAI's deep research. Like the latter, Perplexity Deep Research outperforms several leading LLMs, including OpenAI's o3 mini, DeepSeek R1, Gemini Thinking and Claude 3.5 Sonnet. It should be noted however, that like OpenAI's deep research, Perplexity's product has internet access; something that is not the case for the models the product outperforms. On the SimpleQA benchmark, which tests factual accuracy, Deep Research scored 93.9%, setting a new standard for AI research assistants.
Boldly, Perplexity is making the Deep Research product available to all its customers immediately at launch. Pro subscribers can enjoy unlimited queries, while free users will face an unspecified daily limit. The Deep Research mode is already available in Peplexity website, and will arrive to its iOS, Android and Mac apps shortly.
Ellie Ramirez-Camara is the News Editor at Data Phoenix, where she writes the daily AI newsdesk — covering model releases, research, funding rounds, and policy across the AI and machine-learning industry. She tracks announcements from labs and startups alike and distills them into clear, source-linked reporting for practitioners.
More news

xAI releases Grok Voice Transcribe 2.0 for its speech-to-text API

OpenAI proposes six-pillar youth safety blueprint for Australia

Google confirms Gemini accessed three companies during a safety test
