Perplexity Launches 19-Model AI System for Month-Long Autonomous Tasks

Perplexity Launches Multi-Model AI Agent That Works for Months Without Human Input

Perplexity, the AI-powered search company valued at over $9 billion, launched a platform on Thursday called Perplexity Computer that orchestrates 19 specialized AI models to execute complex, autonomous workflows lasting anywhere from hours to months. The system positions itself not as a chatbot but as a digital employee — one that can manage campaigns, conduct research, file reports, and coordinate between services without waiting for a human to click next.

The launch marks a decisive shift in AI product design. Where most AI tools hand users a response and wait for the next prompt, Perplexity Computer breaks objectives into subtasks, spins up sub-agents for each, integrates external APIs, accesses real file systems, and runs asynchronous execution across isolated computing environments. It has browser access and can interface with third-party software platforms.

Perplexity is making the system available initially to Max subscribers — its highest tier — at a monthly rate that the company has not yet publicly disclosed. The move is part of a deliberate strategy to pivot from ad-supported search revenue toward subscription-based professional tools.

How the Architecture Works

Traditional large language model products rely on a single model generating responses sequentially. Perplexity Computer breaks that mold. A central orchestrator model receives the user's goal — say, "monitor our competitors' pricing changes weekly and report on anomalies" — and decomposes it into discrete subtasks. Specialized models then handle each subtask: one for web search, one for data extraction, one for analysis, one for report writing.

The system can run continuously in the background without any user interaction. If one sub-agent hits an error or an unexpected condition, the orchestrator reassigns the task or flags it for human review rather than failing entirely. Perplexity describes this as "graceful degradation" — a term borrowed from aerospace engineering.

According to Dr. Timnit Gebru, Founder of the Distributed AI Research Institute, "The move toward agentic AI systems capable of operating autonomously for extended periods creates significant questions about accountability, error propagation, and the difficulty of auditing decisions made without human oversight."

Market Implications

The product arrives as Meta, Google, and OpenAI all race to deploy agentic AI systems inside enterprise workflows. Meta embedded its Manus AI agent directly into its Ads Manager platform this week, positioning it to handle campaign research, reporting, and optimization tasks autonomously. OpenAI has been building out its own corporate agent platform with similar ambitions.

For Perplexity, the timing is strategic. The company has faced pressure over its revenue model and its sometimes contentious approach to scraping publisher content for its search product. A pivot to subscription-based professional tools offers a cleaner, more defensible business model — and positions Perplexity in a market that analysts at Goldman Sachs project will exceed $150 billion by 2028.

Whether a 19-model orchestration system can consistently outperform a single, very capable frontier model for complex real-world tasks — or whether the coordination overhead introduces more failure points than it solves — is a question that enterprise customers will answer through adoption over the coming quarters.