Perplexity Ships Portable Computer for Windows on 24GB+ RTX GPUs

Perplexity's on-device agent, Portable Computer, is now available in the Perplexity app for Windows on NVIDIA GeForce RTX and RTX PRO GPUs with 24GB or more of VRAM, NVIDIA wrote on September 14. The Linux build came first, with the Windows release following in September, according to VentureBeat.

The Windows app "runs the model, agent harness, orchestrator, and scheduler entirely on the Windows device," and defaults to "a local model such as Qwen 3.8 27B, post-trained to work with Perplexity Computer and optimized for RTX GPUs." Users can also pick PPLX 27B, "a version Perplexity has post-trained on its own harness," with NVIDIA's Nemotron 3.5 Lightning listed as coming soon. Access is gated to Pro and Max subscribers on individual and enterprise plans, distributed through the Microsoft Store.

The pitch is privacy and price. "Sensitive information stays on device, and locally completed work doesn't consume Perplexity Computer credits," the NVIDIA post states. Connectors for Microsoft Outlook, OneDrive, Word, Google Drive, Gmail, Slack and GitHub let the agent reach across everyday files. For heavier work users can opt in to cloud escalation; VentureBeat reports that a PII classifier previews what would leave the device first, and that using a Claude Opus 5 advisor raised a coding benchmark "from 59.6% to 73.0% at an estimated $0.415 per task."

The 24GB VRAM floor is a real gate. VentureBeat pegs the practical minimum at "roughly a GeForce RTX 3090 or newer," well above what most Windows machines carry. It is one of nearly fifty edge-AI stories we have tracked in the last ninety days, and among the first where a mainstream chatbot vendor treats local execution as the default and cloud as opt-in.