Perplexity Moves Deep Research Into Computer, Routing Research Subtasks Across 20+ Frontier Models For Reports, Decks, And Dashboards

Perplexity has updated its Deep Research feature within Computer, its multi-model orchestration system. This upgrade significantly improves the accuracy, depth of analysis, and citation quality of research outputs. Deep Research now dissects complex questions into smaller subtasks, routing them to over 20 frontier AI models for comprehensive processing.

Deep Research operates by conducting multiple searches, analyzing various sources, and then generating a well-cited report. This enhanced version is integrated into Perplexity Computer, which was launched in February 2026. Computer acts as a cloud-based system capable of coordinating up to 20 AI models within a single workflow. It is designed to be model-agnostic, with Opus 4.6 serving as its primary reasoning engine, and specialized sub-agents like Gemini handling deep research tasks.

The system is built upon two core components: the Agent Search SDK and Search as Code. When presented with a complex inquiry, Deep Research automatically devises a research plan. It then efficiently locates primary sources across hundreds of websites and meticulously cites every piece of information used.

The "Search as Code" functionality enables the system to dynamically construct and execute search processes. This involves running thousands of retrieval steps in parallel, each tailored to the specific question at hand. This adaptive approach, unlike fixed pipelines, allows the system to branch, compare, and refine its search strategy as it learns.

Deep Research in Computer is a premium feature for Perplexity Max subscribers. Developers can access the same advanced research capabilities programmatically through the pay-as-you-go Agent API. The official SDK provides a pre-configured "deep-research" preset to streamline implementation for developers.