At Computex 2026, Perplexity AI unveiled its innovative hybrid local-cloud inference system, claiming it to be the first of its kind that autonomously determines in real-time whether AI workloads should be processed on a user’s device or routed to the cloud. The demonstration, presented by CEO Aravind Srinivas alongside Intel CEO Lip-Bu Tan, highlighted the system’s ability to handle sensitive data locally while allowing more complex tasks to utilize cloud resources. This announcement aligns with the broader trend of on-device AI capabilities shown at Computex, where various chipmakers emphasized hardware designed for local inference. Perplexity’s new offering is poised to strengthen its foothold in enterprise markets, particularly in regulated industries, by ensuring data governance and compliance with sensitive workflows.
Intel: Intel produces processors optimized for client and data center AI workloads. Its Core Ultra Series 3 chips powered the local models in Perplexity’s Computex 2026 demo, enabling hybrid orchestration between device and cloud.
Lip-Bu Tan: Lip-Bu Tan is CEO of Intel. He featured Perplexity’s hybrid inference demonstration in Intel’s Computex 2026 keynote to highlight the company’s client silicon for on-device AI processing.
Perplexity AI: Perplexity AI develops AI agents that orchestrate multiple large language models to handle complex tasks across cloud and local environments. The company demonstrated its hybrid local-cloud inference orchestrator at Computex 2026, where the system autonomously routes sensitive workloads to on-device models and complex reasoning to cloud models. This extends its earlier products like Computer and Personal Computer into dynamic, privacy-focused enterprise workflows.
Aravind Srinivas: Aravind Srinivas serves as CEO of Perplexity AI. He presented the hybrid inference system onstage with Intel leadership at Computex 2026, illustrating real-time decisions on data handling during a demo involving confidential materials.
Enterprise Focus: Perplexity is expanding its agent offerings into regulated industries through enterprise-grade connectors and compliance features that address data governance concerns.
On-device AI Trend: Computex 2026 centered on on-device AI capabilities, with multiple chipmakers showcasing hardware for running inference locally alongside cloud resources.
Competitive Landscape: Major technology companies are developing their own local-cloud AI routing systems, though none currently match the claimed autonomous task-level decision making.