e2b

Sandbox for AI Apps & Agents

Secure sandboxed cloud environments made for AI agents and AI apps

Docs | Website | Discord | Twitter

What is E2B?

E2B Sandbox is a secure sandboxed cloud environment made for AI agents and AI apps. Sandboxes allow AI agents and apps to have long running cloud secure environments. In these environments, large language models can use the same tools as humans do. For example:

• Cloud browsers
• GitHub repositories and CLIs
• Coding tools like linters, autocomplete, "go-to defintion"
• Running LLM generated code
• Audio & video editing

The E2B sandbox can be connected to any LLM and any AI agent or app.

Getting Started & Documentation

Please visit documentation to get started.

To create and control a sandbox, you use our SDK:

Install SDK

npm install e2b

Start sandbox

import { Sandbox } from "e2b";

// Create sandbox
const sandbox = await Sandbox.create();

// Let an LLM use the sandbox here
// Visit https://e2b.dev/docs/sandbox/overview to learn more about sandboxes.

// Close sandbox once done
await sandbox.close();

Development

You can use the SDK with a locally running envd (that usually runs inside the sandbox and allows the SDK to interact with it) by passing E2B_DEBUG=true to .env or by using debug: true in the Sandbox.create or Sandbox.connect method options.

Install dependencies

pnpm install

Generating API clients used by SDK

Check out top-level README.

The API clients are not automatically publicly accessible from the SDK — we are doing some additional work to make the experience of using the APIs better by providing a more user-friendly interface.

Building

For development, you can run the following command to build the SDK on every change:

pnpm dev

Testing

Use pnpm test to run the test suite or pnpm example to run the example code. Pass E2B_DEBUG=true to .env to run against the locally running envd (usually run via Docker).

When running with a local environment the environment is not cleaned up after every test run as when running the test againts production. This might lead to some issues when running the tests multiple times and it is good to keep in mind.