Science & Tech · AI · 5 hrs ago
Perplexity Computer Can Now Run Fully Locally on Macs
Perplexity now lets Mac users run Computer entirely on their device instead of relying on cloud processing for the whole workflow.
The feature works on supported Apple silicon Macs, with 36GB of memory or more recommended, and uses the Qwen 3.6 or PPLX Pro model.
Computer can work with native Mac apps such as Notes, Mail, Messages, Keynote and Numbers, so personal data can stay on the Mac while it handles tasks.
Any code or tool execution happens in isolated sandbox environments.
Users can authorize cloud access when tasks need current information, browser use or cloud-connected apps, and choose whether to use credits or call on more than 15 frontier models for harder problems.
Perplexity’s Hybrid Compute option combines local and cloud models and works on Macs with as little as 24GB of memory.
Local Computer is available now to Pro, Max and Enterprise subscribers through the Perplexity Mac app.
Perplexity says its Computer AI workflow tool can now run fully locally on supported Apple silicon Macs.
The local version uses Qwen 3.6 or Perplexity’s post-trained PPLX Pro model, with 36GB of memory or more recommended.
On-device processing covers the model, task planning, reasoning and file handling, while code and tool execution run in isolated sandboxes.
Users can authorize cloud access for tasks that need current information, browser use or cloud-connected apps, and can call on more than 15 frontier models.
The feature is available to Perplexity Pro, Max and Enterprise subscribers through the Mac app.
- Who
- Perplexity is offering the local Computer feature to its Pro, Max and Enterprise subscribers.
- What
- Perplexity Computer can run fully on a supported Mac, processing its model, planning, reasoning and files on-device.
- When
- October 20, 2026. Perplexity said the feature is available now.
- Where
- On supported Apple silicon Macs, through the Perplexity app for Mac.
- Why
- Perplexity says local processing gives users more control over sensitive data and credit spending.
This story does not have two clearly opposing sides.
Perplexity introduced Hybrid Compute, which splits AI workloads between frontier and local models.
Perplexity announced that Computer could run fully locally on supported Macs.
Perplexity Pro, Max and Enterprise subscribers gained access to the local feature.
- Local models
- Qwen 3.6 or PPLX Pro
- Recommended memory
- 36GB or more for local use
- Hybrid Compute minimum
- 24GB of memory
- Frontier models
- More than 15 can be called on for tougher tasks
- Supported subscribers
- Pro, Max and Enterprise
3 sources
What people are saying
As more demanding workflows become practical to run locally, the machines people already own will take on a larger share of their AI work,
This gives users more control over sensitive data and credit spend.
These same security protocols persist in the cloud version of Perplexity Computer,