Introducing Portable Computer for local-first AI
Portable Computer runs Perplexity Computer entirely on device with NVIDIA, keeping private data local and escalating to the cloud only when a task needs it.
Some of the most valuable work people have for AI agents includes data they’d rather keep on their own machines. Whether it’s private codebases or confidential material, local models are now getting strong enough to work on them right where they are.
Today we’re launching Portable Computer, a version of Perplexity Computer that runs entirely on a local machine. We built it with NVIDIA so users can own the stack on the NVIDIA DGX Spark™, NVIDIA’s desktop-sized AI computer. It will soon also be available on NVIDIA RTX GPU PCs.
Portable Computer lets users run Perplexity Computer entirely on device, analyzing data, synthesizing files, and running complex workflows. Private data stays local and on-device work doesn’t consume credits. When the user authorizes it, the local model can escalate to the cloud for more advanced research and reasoning while still keeping sensitive information on the device.
Local-first AI stack
Portable Computer ships as users seek to get more out of their compute, balancing performance with efficiency. With frontier-like strength available on their own computers for the first time, people no longer have to ration what they use intelligence for.
Portable Computer runs on the NVIDIA DGX Spark with Qwen 3.8 27B or with PPLX 27B, a post-trained version of the Qwen model. NVIDIA Nemotron 3.5 Lightning, a 30B open model, is coming soon to the model picker. The orchestrator, planner, tool router, scheduler, durable task queue, and local search index all run on device.
The Qwen model is post-trained to complete as much of each task locally as possible and to escalate to the cloud when the task needs it. It reads local files, searches across documents and code, takes actions on the device, and keeps jobs running.
Owning the compute makes high-volume AI tasks practical to run because work handled by the local models has no per-credit charge.
Escalation to cloud
Portable Computer handles the majority of tasks anyone needs to run all locally. But some jobs benefit from access to the web or to frontier reasoning in the cloud. A user might want the confidential details of a term sheet to stay local, but escalate to the cloud to get current market comps or recent precedent deals.
Portable Computer’s local orchestrator can escalate a task to the cloud for current information, browser use, connected apps, or one of 15+ frontier models for advanced reasoning.
Through app connectors, Portable Computer works with Google Drive, Gmail, Slack, and GitHub. A user can triage new GitHub issues against overnight bug reports in Gmail on device, then send a Slack message about the top three issues with suggested owners by morning.
When a task needs to send content from the device to a cloud service, Portable Computer asks the user for permission before moving forward.
Search and dictation
A local model only knows what’s on the machine. When a task needs the outside world, Portable Computer can also run Perplexity search or wide or deep research. Users get cited findings from the web without giving up local execution of everything else.
Dictation mode also runs locally with the NVIDIA Nemotron 3.5 ASR Model. Users brief the agent while reading a document or sketching on paper. The transcription and actions on files all stay on the machine. A user can talk through a client situation or dictate over a private codebase without the audio touching the cloud.
Security matches the cloud version of Perplexity Computer. Code and tool execution run in isolated sandbox environments with controlled access to files and connected apps.
Today is just the beginning of local AI
As models get stronger and chips get faster, more people will run complex workflows on their own machines. Every chip cycle and every model release pushes this further. Running AI on personal machines is going to be a much bigger part of how work gets done.
Portable Computer is available to Pro and Max Subscribers on the NVIDIA DGX Spark. The first release is available on Linux, with Windows support coming soon.
The NVIDIA DGX Spark is built around the Grace Blackwell GB10 platform: a 20-core Arm CPU and NVIDIA GPU with 128 GB of unified memory. Portable Computer can be installed with a one-click setup via the Perplexity app.
Read more about the launch on the NVIDIA Local AI blog. Our research blog, A Local-First Agent for Private and Cost-Effective Knowledge Work, details the development behind Portable Computer and its outperformance across benchmarks for accuracy, speed, and credit efficiency.
Try Portable Computer today.