Perplexity unveils hybrid AI for local and cloud models
TL;DR. Perplexity introduced a hybrid AI orchestrator to manage local and cloud models, optimizing privacy and efficiency for its Personal Computer agent. - This system intelligently routes sensitive data to local processing and compute-intensive tasks to the cloud. - The framework will integrate into Perplexity's Always-on agent product, starting in July. - Developed with Intel, the model-agnostic system also supports hardware like Nvidia's RTX Spark.
- Perplexity launched a hybrid inference system that intelligently combines local and cloud AI models.
- The system routes sensitive data locally and computationally intensive tasks to the cloud for optimized privacy and efficiency.
- This technology will be integrated into Perplexity's Always-on agent product, Personal Computer, starting July 2026.
- The model-agnostic framework, developed with Intel, supports various hardware platforms including Nvidia's RTX Spark.