Perplexity has made it possible to run its entire Computer experience, including the harness and orchestrator, directly on a Mac. The feature requires a recent Apple Silicon Mac with at least 36GB of RAM. It builds on Hybrid Compute, an update the company introduced at the start of September that let users split AI workloads between frontier models and local models on their own machine.
Perplexity is pitching the local option as a benefit for budget-conscious users and those concerned about privacy. For example, with access to Notes and Mail, a user could ask Computer to generate a priority to-do list without any email data being sent to the cloud. The company says more demanding workflows will become practical to run locally, allowing the machines people already own to take on a larger share of their AI work. This gives users more control over sensitive data and credit spend.
With today's release, users can choose between two models to run Computer locally: Qwen 3.6 or PPLX Pro. PPLX Pro is a Perplexity offshoot of Qwen 3.6 that the company post-trained to fine-tune it for Computer use. If a task would benefit from a frontier system, users can authorize the software to call on one of more than 15 models. Any code or tool execution is carried out in an isolated sandbox environment, and Perplexity says the same security protocols persist in the cloud version of Perplexity Computer.
Perplexity Pro, Max and Enterprise subscribers can begin using Computer locally starting today. Users need to download and install the latest version of the Perplexity Mac app, then follow the in-app instructions to install a local model on their machine.
Source: Engadget · Summarized by HeadlinesBriefing