Perplexity has launched a new deep-research agent that can autonomously conduct multi-step investigations across the web. Unlike its standard answer engine, the agent plans a research strategy, issues multiple searches, reads sources, and produces a structured report with inline citations.
How the Agent Works
Users enter a research question and set parameters such as depth, source type, and output format. The agent then iterates, refining queries based on intermediate findings. It can compare competing claims, identify gaps in available information, and flag uncertain conclusions. The final report includes a bibliography and transparent source links.
Behind the scenes, the agent uses a planning model to decide which sources to consult and how to structure the investigation. It can also generate sub-questions, search academic repositories, and cross-check facts across independent outlets.
Use Cases and Accuracy
Perplexity is targeting analysts, journalists, academics, and consultants who need quick but well-sourced background research. The company acknowledges that the agent can still hallucinate or misinterpret sources, and it recommends human verification for high-stakes decisions. A confidence score and uncertainty flagging are built into the interface.
Competitive Landscape
The launch puts Perplexity in closer competition with OpenAI’s research previews, Google’s AI Overviews, and dedicated research tools like Elicit and Consensus. Perplexity’s differentiator remains its clean interface and source-first approach. The feature is available on Pro and Enterprise plans.
Industry Impact
Industry watchers view this announcement as another sign that the artificial intelligence market is shifting from raw capability demonstrations toward production-ready features. Buyers are increasingly focused on total cost of ownership, data governance, vendor transparency, and long-term support. The move also pressures competitors to respond quickly, which should accelerate innovation and drive more flexible pricing across the market. For end users, the practical result is likely to be better tools, clearer licensing terms, and stronger safety guardrails as the industry matures through 2025 and 2026. Enterprises that move early may capture meaningful workflow efficiencies before these capabilities become table stakes.