Perplexity AI Updates: June 7, 2026

1. Perplexity Unveils Hybrid Local-Server Inference Orchestrator

Perplexity. Perplexity introduced what it calls the first hybrid local-server inference orchestrator for its Personal Computer product, demonstrated at Computex 2026. A compact on-device model evaluates each task and automatically routes sensitive or lightweight work to local models while sending complex workloads to frontier cloud models, coordinating up to 20 models across execution locations. The framework is model and hardware agnostic, with confirmed support for Intel Core Ultra Series 3 and NVIDIA RTX Spark, and is set to arrive in July 2026. Source