Challenges With Perplexity Portable Computer

Perplexity’s new local-first AI agent aims to address pressing concerns in the AI market: high token costs and the need for privacy. While the agent is part of the growing trend of edge AI, it also presents challenges, particularly the high cost of the AI computer it requires.The AI search vendor on Tuesday introduced Portable Computer, a version of Perplexity Computer that runs locally. First introduced in February, Computer is a general-purpose digital worker that assists different AI models in performing autonomous workflows.With Portable Computer, users run Computer on their personal device, particularly Nvidia DGX Spark, a compact desktop supercomputer. Perplexity built the local agent with Nvidia, so it runs on the personal AI computer with the Qwen 3.8 27B model from Chinese tech giant Alibaba, or PPLX 27, a post-trained version of Qwen.Portable Computer is an example of the movement toward more agentic AI running on edge devices to reduce token costs and protect user privacy. Since the release of OpenClaw in 2025, vendors have been focusing on models or agents that can run on users’ devices. In June, Microsoft introduced its family of Aion models, which are designed to orchestrate local agentic workflows. In March, Anthropic updated its Claude model to have access to local computers, and open source AI lab Ai2 introduced a computer use agent for users’ desktops.Related:Mistral and Saudi Vendor to Advance Sovereign AI in Middle EastEdge AI“[Portable Computer] is another good example of the evolution toward local AI, enterprise AI being more important, private AI, people really caring about where does my AI run, what data can it access,” said Said Oouisaal, CEO and founder of edge computing vendor Zendeda. “You want to own your intelligence, and this is a good example of that.”Perplexity said confidential data stays local and on-device and does not use many tokens or credits. The model manages most tasks locally on the computer. Still, some tasks are escalated to the cloud if needed for browser use, connected apps or other frontier models for advanced reasoning.While this helps users looking to own their AI, it also benefits Perplexity, in that it will use Portable Computer as an opportunity to help the vendor appeal to new customers, said Lian Jye Su, an analyst at Omdia, a division of Informa TechTarget.“It allows them to expand their customer reach a little bit,” Su said. “[They’re] using this as a gateway for people who are more cost-conscious or have a lot of sensitive documents that they are not willing to upload to the cloud.”Related:Alibaba Sells More Shares to Raise $10.2B to Spend on AIAnd for Nvidia, which helped build the local agent, it is a means to expand the use of its DGX devices.Some ProblemsWhile Portable Computer lets users be more cost-conscious with token spending, it still has limitations. For example, users save money on tokens but incur significant costs with DGX Spark. A single DGX Spark computer costs $4,699.“It’s not cheap,” Su said. “To buy DGX is not a random thing. It’s a very deliberate decision, and I do not think anyone is doing that lightly … even enterprises.”If Perplexity is intent on selling an approachable technology that enterprises can use, it likely needs to find a way to work with Microsoft and the Windows ecosystem, he said.Another challenge is the difficulty users encounter when working with an AI computer.“It’s still quite complex to set up a system to do inference on your own,” Ouisaal said. He added that while it’s easy to buy the hardware, it’s harder to install the layers of software needed, such as Linux or container software such as Kubernetes, so that locally created and run AI agents can function properly. He added that while t Portable Computer is a single-function system, in an enterprise environment, more expertise will be needed to serve different departments and agents.Related:Thomson Reuters’ New Model Could Inspire Other SaaS Vendors“It’s not easy,” Ouisaal continued. “You need to be an engineer.”However, it is unclear whether Portable Computer will ever make its way into the enterprise market in significant numbers in its current iteration, Su said.“Enterprises are already really concerned with shadow AI and are already restricting machines that have internet access to all these [models],” he said. He added that while Perplexity and Nvidia can claim that only token-consumption data is transferred to their servers and all other data stays on-device, enterprises might still be concerned about giving employees more independent access to AI.About the AuthorNews Writer, AI BusinessEsther Shittu has covered AI technologies and industry trends since 2021. As co-host of the Targeting AI podcast, she talks with experts, thought leaders and practitioners exploring critical AI developments. Before AI Business, she wrote for SearchEnterpriseAI, the New York Daily News, Bklyner and the Brooklyn Daily Eagle. When she's not diving deep into the world of AI, she spends her time on passion projects and raising her three daughters.