Perplexity introduces Computer, a system that creates and executes complete workflows through multi-model orchestration. The system coordinates multiple sub-agents that perform tasks related to research, coding, and document generation. It can run autonomously for hours to months.
Where chat interfaces provide answers and agents perform tasks, Computer creates complete workflows. Users describe the desired result, after which the system automatically breaks it down into tasks and subtasks.
The subagents work asynchronously and can perform web research, draft documents, process data, or make API calls to connected services. For example, one agent can draft a document while another simultaneously collects the necessary data. When Computer encounters a problem, it independently creates subagents to solve it.
Perplexity Computer runs on Opus 4.6 as its core reasoning engine and automatically selects the best model for specific subtasks. Gemini is used for deep research and creating subagents, Nano Banana for images, Veo 3.1 for video, Grok for fast tasks, and ChatGPT 5.2 for long-context recall.
From search to agentic systems
Perplexity previously developed Comet, an AI-native browser, and Comet Assistant as a personal AI agent. Those foundations led to the current focus on multi-agent orchestration. Perplexity emphasizes that the system remains model-agnostic, giving users choice and control. Because token budgets play a more important role in different functions, users can choose specific models for specific subtasks.
Each task runs in an isolated compute environment with access to a file system, browser, and tool integrations. According to Perplexity, the result is a secure environment for powerful AI, universally accessible without local installation.
Perplexity Computer is available to Perplexity Max subscribers and will soon be available to Enterprise Max users as well.
Tip: Perplexity expands cloud strategy with Azure alongside AWS