You'll need a Mac with at least 36GB of RAM, however.
Perplexity
At the start of September, Perplexity introduced Hybrid Compute, an update to its Computer app that gave users the option to split their AI workloads across a combination of frontier and local models running on their own machine. In hindsight, it was probably only a matter of time before the company made it possible for people to run the platform entirely on their own device, because that's exactly what it's doing now.
Provided you own a recent Apple Silicon Mac with at least 36GB of RAM, now you can run the entire Computer experience, including the all-important harness and orchestrator, on your own machine. Like Hybrid Compute, Perplexity is pitching this feature as a boon to thrifty users and those who want to keep their data private. For example, with access to your Notes and Mail apps, you could ask Computer to generate a priority to-do list for you, without any of the data in your emails being sent to the cloud. "As more demanding workflows become practical to run locally, the machines people already own will take on a larger share of their AI work," the company explains. "This gives users more control over sensitive data and credit spend."
With today's release, you can choose one of two models to run Computer locally: Qwen 3.6 or PPLX Pro. The latter is a Perplexity offshoot of Qwen 3.6, which the company did its own post-training for to fine tune it for Computer use. If, at any point, you feel like the task you're running would benefit from a frontier system chipping in, you can authorize the software to call on one of more than 15 models. Any code or tool execution a task might require is done in an isolated sandbox environment. "These same security protocols persist in the cloud version of Perplexity Computer," says Perplexity.
Perplexity Pro, Max and Enterprise subscribers begin using Computer locally starting today. Once you download and install the latest version of the Perplexity Mac app, just follow the in-app instructions to download and install a local model on your machine.