# Perplexity's local agent reaches Windows machines as research tackles the context budget

Portable Computer runs the model, harness and scheduler on-device on RTX GPUs with 24 gigabytes of memory or more, while new work proposes a tiered protocol for measuring what memory strategies cost local agents.

- Published: 2026-09-15T06:27:29.669Z
- Canonical: https://polylog.news/ai/2026-09-15/perplexity-s-local-agent-reaches-windows-machines-as-researc
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [NVIDIA Blog](https://blogs.nvidia.com/blog/local-ai-perplexity-windows-pcs/), [arXiv (BudgetBench)](https://arxiv.org/abs/2609.13149), [arXiv (Lexical Prompt Compression)](https://arxiv.org/abs/2609.13154), [OpenAI](https://openai.com/index/perplexity-improving-accuracy-with-astra)

Perplexity brought Portable Computer, the on-device version of its Perplexity Computer agent, to Windows on Sunday, running on GeForce RTX and RTX PRO hardware with at least 24 gigabytes of graphics memory, according to NVIDIA. The full loo…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-09-15/perplexity-s-local-agent-reaches-windows-machines-as-researc (subscription information: https://polylog.news/pricing).