Zo Computer Features and How It Works

Zo Computer is a personal AI cloud computer that gives you a persistent Linux server, 100GB of cloud storage, and an always-on AI. This article covers every major feature in detail, what it does, how it works, and what you can build with it.

If you are new to Zo, start with What Is Zo Computer for the overview. This article goes deeper on the technical capabilities.

The persistent Linux server

At the core of Zo is a Linux server running in the cloud. It belongs to you. It stays on.

This is the feature that makes everything else possible. Most AI tools reset at the end of a conversation. Zo’s server persists across sessions, which means long-running processes, hosted services, and scheduled tasks all continue whether or not you are actively using the product.

• Always-on compute. On paid plans, the server never goes to sleep. Background tasks keep running.
• Shell access. The AI can execute shell commands directly on your server.
• Your own environment. Files, installed packages, and running processes are yours. Nothing is shared with other users.

The free plan has one meaningful constraint: the server goes to sleep when inactive. For any workflow that needs to stay live, the Basic plan at $18/month removes that limitation.

File storage and the filesystem

Zo gives you 100GB of cloud storage with a full Linux filesystem.

This is not a limited document vault. It is a real filesystem. You can store:

• Documents and notes. Text files, PDFs, spreadsheets, presentations.
• Code repositories. Clone a repo, edit it, run it, all from within your Zo environment.
• Images and media. Generated images, audio files, project assets.
• Project folders. Organize your work the same way you would on a local machine.

The AI can read from and write to this filesystem natively. You do not need to copy and paste content into a chat window. You can tell Zo to open a file, edit a section, and save the result.

Having your files and your AI in the same environment is a fundamentally different experience from uploading attachments to a chat session every time you need help with something.

AI conversation and model access

Zo gives you access to the latest AI models for chatting, generating media, and transcribing audio.

You can interact with Zo via a browser interface or by text from your phone. The AI is not a single fixed model. Zo provides access to current models, which means you are not locked into a version that was current a year ago.

Capabilities include:

• Chat and reasoning. Ask questions, work through problems, get analysis on documents in your filesystem.
• Media generation. Generate images directly from a conversation.
• Audio transcription. Upload or stream audio and get a transcription.
• Web search. Zo can search the web in real time and bring results into your conversation.
• News. Access current news as part of any task or scheduled briefing.

Task scheduling — AI that runs without you

Task scheduling is one of Zo’s most powerful and underused features. You can define a recurring AI task and have it run automatically, with results delivered via email or SMS.

The types of tasks you can schedule include:

• Check email. Zo reviews your inbox on a schedule and sends you a prioritized summary.
• Generate reports. Pull data from a connected app, format a report, and email it to stakeholders.
• Monitor websites. Check a page for changes, competitor pricing, job listings, regulatory updates, and alert you when something shifts.
• Post to social media. Draft and publish content on a recurring schedule.

The delivery options are practical. Results come to you via email or SMS, which means the output finds you rather than requiring you to log in to check.

No external automation tool is required. No Zapier, no Make, no n8n. You describe the task and the schedule to Zo and it handles execution.

Website and app hosting

Zo supports production-grade hosting for websites, web applications, backend services, and databases.

• React apps. Deploy a full front-end application directly from your Zo environment.
• API servers. Host a backend API with a permanent URL. Always on, always accessible.
• Databases. Run a database alongside your application on the same server.
• HTTPS and custom domains. All hosted projects get HTTPS by default. You can point a custom domain to any project.

The practical effect is that your development environment and your deployment environment are the same place. You edit a file in your Zo filesystem, and the change is live. There is no separate hosting bill and no separate deployment pipeline to maintain for personal or small-team projects.

Native code execution

Zo supports native code execution in a secure sandboxed environment. You can run Python or JavaScript directly.

This matters for a specific set of tasks:

• Data processing. Run a Python script on a dataset stored in your filesystem.
• Automation scripts. Execute JavaScript to interact with APIs or transform data.
• Testing. Run code before committing it or deploying it.

The sandbox means execution is isolated. Your server and files are not exposed by running untrusted code. You get real compute without real risk.

The MCP server — connecting your dev tools

Zo includes an MCP server that connects developer tools directly to your personal cloud.

Supported tools include:

• Claude Code
• Cursor
• Gemini CLI
• Codex

When you connect one of these tools to your Zo via MCP, your IDE gets access to your cloud files, your connected integrations, and all of Zo’s built-in tools. This means you can:

• Read and write files on your Zo filesystem from inside your code editor
• Use Zo’s integrations (Gmail, Notion, Airtable) from within your IDE
• Run shell commands on your cloud server from a local development session
• Access Zo’s full 50+ tool suite without leaving your editor

For developers who use Claude Code regularly, this is the most direct integration available. See Zo Computer for developers: MCP and API guide for setup details.

The MCP server is the bridge between your local development workflow and your personal cloud. It makes Zo feel like a native extension of your IDE rather than a separate product to manage.

Integrations

Zo connects to a range of external services. These integrations are available to the AI during conversations, in scheduled tasks, and via the MCP server.

These integrations are most powerful when combined with task scheduling. A daily briefing that checks Gmail, pulls open Linear issues, and delivers a formatted summary via SMS is a single scheduled task, not a multi-tool automation workflow.

The Skills registry and built-in tools

Zo includes 50+ built-in tools accessible through the Skills registry at zo.computer/tools.

Built-in tools cover a wide range of capabilities:

• File read, write, and search. Work with your filesystem without manual navigation.
• Shell commands. Run terminal commands via the AI interface.
• Web browsing. The AI can open URLs, extract content, and interact with web pages.
• Image generation. Generate images from text descriptions in any conversation.
• Web search and code search. Find information or locate code across the web.
• News. Pull current news on any topic as part of a task or conversation.
• Texting and emailing. Send messages directly from Zo as part of any workflow.

The Skills registry is where you can see and manage which tools are active in your environment. It is also where AI agents draw their capabilities from when executing tasks on your behalf.

AI agents

Zo supports AI agents, autonomous processes that can take a sequence of actions to complete a multi-step task.

An agent on Zo has access to your files, your integrations, your scheduled tasks, and all 50+ built-in tools. This is different from a simple AI assistant that can only respond to the current message. An agent can plan a sequence of steps, execute them, and report back with results.

This capability is what makes complex scheduled workflows possible. Zo is not just running a script on a timer. It is running an AI that can adapt to what it finds when it checks your inbox or monitors a website.

What Zo works well for

Zo is the right tool for users who need AI to do more than answer questions. Specific use cases include:

• Developers who want to connect Claude Code or Cursor to a personal cloud via MCP
• Power users who want recurring AI tasks delivered without manual effort
• Small teams who need a simple, always-on AI infrastructure without a complex stack
• Anyone who wants AI with persistent memory and real compute behind it

For broader context on use cases, see Zo Computer use cases.

Common questions on Zo Computer features

”Can Zo run tasks when I’m not logged in?”

Yes. On the Basic and Ultra plans, the server is always on. Scheduled tasks run on their defined schedule regardless of whether you are actively using Zo.

”Is native code execution safe?”

Yes. Python and JavaScript run in a secure sandboxed environment. Execution is isolated from your main server environment, which means running untrusted code does not expose your files or running processes.

”How do I connect Claude Code to Zo?”

Zo provides an MCP server that Claude Code connects to via standard MCP configuration. Once connected, Claude Code has access to your Zo filesystem, integrations, and built-in tools. See zo.computer/tools and the documentation at zocomputer.mintlify.app for setup steps.

”Can I bring my own API keys?”

Yes. All Zo plans support bringing your own API keys. AI usage beyond included credits is charged at cost, Zo does not mark up model usage.

Get Zo working for your workflow

Zo’s feature set is broad enough that knowing where to start matters. The most common first step is connecting your most-used integration and defining one recurring task.

The users who get the most from Zo are the ones who give it something to run automatically — not just something to respond to.

Path one: start with the free plan. Go to zo.computer, explore the Skills registry at zo.computer/tools, connect Gmail or Notion, and schedule one recurring task this week. See what shows up in your inbox on day two.

Path two: bring in a partner. Phos AI Labs helps teams identify the highest-value workflows to build on personal AI infrastructure like Zo. Thirty minutes, no deck. Start here.