Perplexity launches Personal Computer AI tool for Windows, intensifying the race for your desktop

Via cnet.com

Perplexity launches Personal Computer AI tool for Windows, intensifying the race for your desktop

The AI search company's agentic tool now reaches over 1 billion Windows devices, marking another front in the battle between AI platforms and crypto-native alternatives for user attention and data.

Perplexity just brought its AI agent to the world’s most popular operating system. The company announced on June 3 that its “Personal Computer” tool, previously available only on Mac since April, now supports Windows PCs through a waitlist for Max and Enterprise Max subscribers.

The tool functions as what Perplexity calls a “general-purpose digital worker.” It can access your local files, open your apps, create documents, and update spreadsheets. It also integrates with Microsoft 365 and Teams, integrations that launched in May.

What Perplexity is actually building

Personal Computer for Windows builds on a broader “Computer” agent platform that Perplexity debuted in February 2026. Windows powers over 1 billion devices globally.

Advertisement

At Computex 2026, the company demoed a hybrid local-cloud inference feature that processes some AI tasks directly on your machine rather than routing everything through remote servers. That capability is slated for an initial rollout in July 2026, and it will be exclusive to Windows first.

Perplexity reportedly saw approximately 50% revenue growth in the month following the February 2026 launch of its Computer agent platform.

The AI-crypto intersection nobody’s talking about

The race to build agentic AI is happening on two parallel tracks. One track runs through companies like Perplexity, OpenAI, and Google, building centralized AI agents tied to traditional subscription models. The other track runs through crypto-native projects building decentralized AI agents using token incentives and on-chain infrastructure.

Perplexity’s hybrid local-cloud approach reduces cloud compute costs by offloading work to local hardware. Decentralized compute networks that sell GPU cycles need centralized AI companies to remain dependent on cloud infrastructure. Every efficiency gain on the centralized side puts pressure on the demand thesis for decentralized alternatives.

What this means for investors

The reported 50% revenue bump following the Computer platform launch suggests strong product-market fit. If the Windows expansion delivers similar results given the much larger addressable market of over 1 billion devices, Perplexity could establish itself as the default AI agent layer for enterprise workflows.

The key metric to watch is whether Perplexity’s July rollout of local inference on Windows changes the compute demand picture. If enterprises start running AI workloads on their own hardware at scale, that could shift spending away from both cloud providers and decentralized compute networks simultaneously.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

Perplexity launches Personal Computer AI tool for Windows, intensifying the race for your desktop

Perplexity launches Personal Computer AI tool for Windows, intensifying the race for your desktop

The AI search company's agentic tool now reaches over 1 billion Windows devices, marking another front in the battle between AI platforms and crypto-native alternatives for user attention and data.

Via cnet.com

Perplexity just brought its AI agent to the world’s most popular operating system. The company announced on June 3 that its “Personal Computer” tool, previously available only on Mac since April, now supports Windows PCs through a waitlist for Max and Enterprise Max subscribers.

The tool functions as what Perplexity calls a “general-purpose digital worker.” It can access your local files, open your apps, create documents, and update spreadsheets. It also integrates with Microsoft 365 and Teams, integrations that launched in May.

What Perplexity is actually building

Personal Computer for Windows builds on a broader “Computer” agent platform that Perplexity debuted in February 2026. Windows powers over 1 billion devices globally.

Advertisement

At Computex 2026, the company demoed a hybrid local-cloud inference feature that processes some AI tasks directly on your machine rather than routing everything through remote servers. That capability is slated for an initial rollout in July 2026, and it will be exclusive to Windows first.

Perplexity reportedly saw approximately 50% revenue growth in the month following the February 2026 launch of its Computer agent platform.

The AI-crypto intersection nobody’s talking about

The race to build agentic AI is happening on two parallel tracks. One track runs through companies like Perplexity, OpenAI, and Google, building centralized AI agents tied to traditional subscription models. The other track runs through crypto-native projects building decentralized AI agents using token incentives and on-chain infrastructure.

Perplexity’s hybrid local-cloud approach reduces cloud compute costs by offloading work to local hardware. Decentralized compute networks that sell GPU cycles need centralized AI companies to remain dependent on cloud infrastructure. Every efficiency gain on the centralized side puts pressure on the demand thesis for decentralized alternatives.

What this means for investors

The reported 50% revenue bump following the Computer platform launch suggests strong product-market fit. If the Windows expansion delivers similar results given the much larger addressable market of over 1 billion devices, Perplexity could establish itself as the default AI agent layer for enterprise workflows.

The key metric to watch is whether Perplexity’s July rollout of local inference on Windows changes the compute demand picture. If enterprises start running AI workloads on their own hardware at scale, that could shift spending away from both cloud providers and decentralized compute networks simultaneously.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.