Perplexity AI Unveils 'Perplexity Computer'
Perplexity AI has launched "Perplexity Computer," a new agentic AI tool designed to execute complex, multi-step tasks. The company demonstrated its capability by building a live web application that processes and displays satellite orbital data.
This launch marks a significant strategic pivot for Perplexity AI, moving it beyond its identity as an "answer engine" and into the rapidly growing agentic AI market. The company, founded in 2022, is now leveraging its $20 billion valuation to transition from providing cited answers to autonomously executing complex tasks. "Perplexity Computer" operates by orchestrating 19 different specialized AI models, including those from competitors like Google (Gemini) and Anthropic (Claude). The system uses a planner-executor model to break down high-level goals into sub-tasks, assigning each to the most suitable AI for the job, such as deep research, coding, or data analysis. This multi-model approach is CEO Aravind Srinivas's bet that the future of AI lies not in a single master model, but in the effective coordination of many specialized ones. Enterprise usage data from Perplexity showed that by late 2025, no single AI model held more than a 25% share for tasks, indicating a shift towards a need for orchestration. The move positions Perplexity to compete in a market for AI agents that is projected to expand significantly. This broader industry trend is shifting focus from "human-in-the-loop" copilots to "human-on-the-loop" autonomous systems that can manage entire workflows. Some analysts see this as a direct challenge to the API-based agent offerings from major players like OpenAI and Google. This new tool reflects a larger industry race to create a "general-purpose digital worker." The goal is to evolve AI from an assistant that provides information into an "action engine" that delivers completed projects, a transition Perplexity is making to move up the value chain beyond AI-powered search. The introduction of "Perplexity Computer" also signals a shift in business models, with a focus on usage-based pricing for completing tasks rather than just monthly subscriptions for access. This aligns with the idea of AI agents as a new form of labor that companies can hire to perform digital work.
