Perplexity Computer runs fully offline on Apple Silicon Macs using Qwen 3.6 or PPLX Pro, with 36GB RAM recommended
The Verge reported Oct 20 that Perplexity Computer can now run fully offline on Apple Silicon Macs, handling tasks across apps and files without sending data to the cloud, running Qwen 3.6 or PPLX Pro locally, with 36GB of RAM or more recommended.
The Verge reported on Oct 20 that Perplexity Computer can now run fully offline on Mac. The update lets users perform tasks across their apps and files without transmitting data to the cloud, and Perplexity recommends 36GB of RAM or more on Apple Silicon.
On the model side, the product runs Qwen 3.6 or PPLX Pro directly on the Mac, with PPLX Pro being a post-trained version of Qwen. The distinction matters when weighing subscription value: with local inference data can stay on the device, whereas offloading a task to the cloud involves upload.
On the feature boundary, the report notes users can still transfer certain tasks to the cloud for access to updated information and browser use. In other words, fully offline mainly means inference and task handling can happen locally with local task data not sent to the cloud; it does not mean every capability is available offline, and tasks needing current information or browser access can still be sent to the cloud.
The report does not list specific supported models such as MacBook Air, MacBook Pro, Mac mini or Mac Studio, nor the minimum supported Apple Silicon chip or macOS version, so the accurate framing is Apple Silicon Macs with 36GB or more of unified memory. It provides no data on local speed, supported apps, performance or the specific privacy mechanism.