Perplexity announces hybrid AI system that decides what runs locally or in the cloud
Quick Answer
Perplexity has unveiled a hybrid AI orchestrator that intelligently allocates tasks between local and cloud-based models, optimizing processing efficiency.
Quick Take
Perplexity has unveiled a hybrid AI orchestrator that intelligently allocates tasks between local and cloud-based models, optimizing processing efficiency. This system aims to enhance user experience by leveraging the strengths of both local computing power and cloud capabilities, ensuring that tasks are executed in the most effective environment.
Key Points
- The orchestrator automatically determines whether tasks run locally or in the cloud.
- Combines local AI models with powerful cloud-based models for optimal performance.
- Aims to improve processing efficiency and user experience across applications.
- Targets users needing flexible AI solutions that leverage both local and cloud resources.
📖 Reader Mode
~1 min readPerplexity has announced an orchestrator that combines AI models running on your own computer with powerful cloud models and automatically decides which task gets processed where. The goal is to optimize accuracy, privacy, and energy efficiency at the same time. The hybrid inference system will be integrated into the Always-on agent product Personal Computer, which was introduced in March, starting in July.
Sensitive data like financial documents or health information will stay local, while compute-intensive tasks get routed to cloud models. Perplexity introduced the system together with Intel. But the model-agnostic framework also runs on other hardware, like Nvidia's RTX Spark. "The race for local compute is on," the announcement says.
According to Perplexity, shifting routine tasks to local devices could reduce the need for centralized computing infrastructure and simplify questions of data sovereignty. The company says its business model rewards correct answers instead of high compute consumption—which makes optimizing for efficiency a natural incentive.
— Originally published at the-decoder.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.
