# E2B

> Open-source cloud infrastructure providing secure sandboxes for AI agents to run code in isolated environments. Sandboxes start in 80ms and support Python, JavaScript, Ruby, and C++ on Linux.

- **URL:** https://e2b.dev/
- **Category:** LLM Tools
- **Tags:** sandbox, security, infrastructure
- **Updated:** 2026-03-24 13:36
- **Canonical page:** https://vectordb.ever.works/items/e2b

## Details

## Overview

E2B is an open-source infrastructure that provides secure cloud sandboxes for AI agents, allowing them to run AI-generated code in isolated environments. The platform can run Python, JavaScript, Ruby, or C++ code on Linux.

## Key Features

**Performance**: E2B Sandboxes in the same region as the client start in 80 ms, and each sandbox can run up to 24 hours.

**Security**: Each sandbox is powered by Firecracker, a microVM made to run untrusted workflows.

**Compatibility**: E2B is LLM-agnostic and compatible with any model, supporting OpenAI, Llama, Anthropic, Mistral, and custom models.

## Recent Developments (2026)

**Docker Partnership**: In December 2025, E2B announced a partnership with Docker, and now every E2B sandbox includes direct access to Docker's MCP Catalog, a collection of 200+ tools.

**Use Cases**: E2B enables developers to build AI agents that see, understand, and control virtual Linux desktops using E2B Desktop sandboxes.

The platform is actively maintained, with the latest SDK release on March 18, 2026.

## Integration

E2B integrates with popular frameworks:
- smolagents framework supports E2B sandboxes
- LangChain, LlamaIndex, and other frameworks
- Works with all major LLMs

## Pricing

- Free tier available
- Pay-as-you-go pricing
- Enterprise plans available

---

_This Markdown mirror is generated by Ever Works for AI agents. The canonical HTML page is at https://vectordb.ever.works/items/e2b._