Change language to
0:00

Perplexity Portable Computer is now available in the Perplexity app for Windows PCs with compatible NVIDIA RTX or RTX PRO GPUs, bringing local AI workflows to a wider range of machines. Perplexity Portable Computer keeps its model, agent harness, orchestrator and scheduler on the Windows device for on-device work, while allowing users to approve cloud escalation for tasks that need stronger reasoning or current web information.

The Windows announcement comes from Perplexity, while NVIDIA’s account details the RTX hardware requirement and local setup.

Blurred person carrying a laptop passes a desktop monitor displaying an AI assistant interface in a warmly lit office

Perplexity Portable Computer reaches Windows RTX PCs

The Windows release is available to Pro and Max subscribers on individual and enterprise plans. On-device inference requires an NVIDIA GeForce RTX or RTX PRO GPU with at least 24GB of VRAM, according to Perplexity. That rules out most mainstream gaming laptops and desktops, including many newer cards with 16GB of memory.

Local work does not consume Perplexity Computer credits. The software can analyse files, combine information across a device, connect to tools through local Model Context Protocol servers and run scheduled tasks. Perplexity says the Windows app can work with Gmail, Outlook, Slack and GitHub, while code and tool execution run in isolated sandbox environments.

What runs locally, and what still goes to the cloud?

The local setup is designed for routine file handling and recurring workflows. When a task needs web access, advanced reasoning or another cloud capability, Portable Computer can send the relevant step to Perplexity’s online models after asking for permission. Perplexity says its service can access more than 15 cloud models for those escalations.

The Windows launch follows Portable Computer’s August debut on NVIDIA DGX Spark, where Perplexity offered PPLX 27B and Qwen 3.8 27B. Perplexity’s current Windows documentation lists PPLX 27B for RTX PCs, while NVIDIA’s announcement describes the local model as a post-trained Qwen 3.8 27B model optimised for its agent workflows. The two descriptions refer to the same model family, but the Windows product page is the more specific source for the supported setup.

The release is available through the existing Perplexity app for Windows. There is no UAE-specific pricing or availability information in the announcement, and the 24GB VRAM requirement means local access will initially appeal to owners of high-end workstations and enthusiast PCs rather than ordinary office machines.

NEWSLETTERS

Subscribe to our Newsletters

Two newsletters. Zero noise. Pick what lands in your inbox.

Unsubscribe anytime. We don’t share your email.

What is Perplexity Portable Computer?

It is Perplexity’s local AI agent platform for planning and carrying out multi-step tasks on supported hardware.

What GPU does Portable Computer for Windows require?

Perplexity says Windows on-device inference requires a compatible NVIDIA GeForce RTX or RTX PRO GPU with at least 24GB of VRAM.

Does local work use Perplexity Computer credits?

Work completed on the Windows device does not consume Perplexity Computer credits, but approved cloud steps can still use online services.