Cloudflare Containers.
Ready for your agent.

The hosted SaaS layer on top of Cloudflare Containers. Start Linux machines by API, run commands, upload files, SSH in, and open web previews. From $5/month.

No Cloudflare account or Worker deployment needed.

Agent workspaceJavaScript SDK
import { Mainbrella } from '@mainbrella/sdk';
const client = new Mainbrella({
  apiKey: process.env.MAINBRELLA_API_KEY
});
const machine = await client.create({
  catalogId: 'node', size: 'small'
});
try {
  const result = await machine.commands.run(
    'node --version'
  );
  console.log(result.stdout);
} finally { await machine.kill(); }

Use the computer. Skip building the control plane.

Cloudflare supplies the runtime. Mainbrella supplies the service.

Cloudflare Containers is infrastructure for running container workloads through Workers and Durable Objects. Mainbrella adds customer accounts, billing, named API keys, machine lifecycle controls, command and file APIs, browser terminals, SSH, and previews.

A hosted Mainbrella account does not require your own Cloudflare account or Worker deployment. If you need direct control of your Worker architecture, you can use Cloudflare Containers directly or self-host Mainbrella.

Cloudflare Containers documentation · Mainbrella is an independent service, not a Cloudflare product.

Before you start

Do I need a Cloudflare account?

Not for Mainbrella Hosted. Sign in, choose a plan, create an API key, and use the managed service. Self-hosting requires your own Cloudflare infrastructure.

How are environments isolated?

Cloudflare runs each container in a separate microVM. Mainbrella checks account ownership and container generations at the API boundary. Read the security details for the service’s controls and limitations.

How does the $5 minimum work?

A $5 one-time top-up adds $5 of prepaid credit. Compute costs $0.02 per unit-hour, subject to your balance and separate monthly spending cap. Prepaid accounts support 10,000 starts per UTC month and 100 containers within 128 concurrent units. Sessions last up to 24 hours with a 30-minute idle timeout. Add balance for more compute.

Can I run a permanent server here?

Mainbrella’s current machines are temporary, with idle and hard session deadlines. Save a filesystem snapshot or export your files before stopping. For permanent services, check those lifecycle requirements first.

Try your actual workload.

Start with usage billing, create an API key, and run your first machine. Need help checking fit? Talk to Andrew.

Start building — $5 minimum

Plan limits · Security · Service status · Open source

Works with your agent.

Use skill instructions with a coding agent, or connect HTTP tools in your application.

Five-step setup instructions

Step 1: Set up Mainbrella

Read https://mainbrella.com/SKILL.md and follow its Quick Setup. It covers the API key handoff (MAINBRELLA_API_KEY, starts with mb_), prerequisites, and the read-only doctor command. Ask the user only where the skill says to.

Step 2: Detect the project

Do not ask. package.json means JavaScript/TypeScript, pyproject.toml or requirements.txt means Python, Cargo.toml means Rust, and go.mod means Go. An empty directory gets a minimal Python project in a new subdirectory unless the user named a language.

Step 3: Integrate

Follow the skill's HTTP workflow, matching the project's package manager, module format, and environment loading. Select an available image for the project. If agent or tool-calling code already exists, wire Mainbrella in where command execution happens and keep the existing conventions.

Step 4: Verify

Run the skill's verification script: create a container, run echo "hello from mainbrella", check stdout and exit code, and stop the created container in a finally block. Pass means stdout "hello from mainbrella", exit code 0, and cleanup completed. If it fails, follow Troubleshooting and re-run once only after cleanup is confirmed and allowance permits. Preserve existing containers.

Step 5: Report

One paragraph: what changed, what ran, and what you verified. Then offer one next step from the skill: a custom image, an LLM tool that runs commands, or a coding agent running inside a container.