How to beginner · 3 min read

How to install packages in E2B sandbox

Q: How to install packages in E2B sandbox

To install packages in the E2B sandbox, run pip install commands inside the sandbox using Sandbox.run_code(). This executes installation commands securely within the sandbox environment without affecting your local system.

Quick answer

To install packages in the E2B sandbox, run pip install commands inside the sandbox using Sandbox.run_code(). This executes installation commands securely within the sandbox environment without affecting your local system.

PREREQUISITES

Python 3.8+
E2B API key
pip install e2b-code-interpreter

Setup

Install the e2b-code-interpreter Python package and set your E2B_API_KEY environment variable before running code in the sandbox.

bash

pip install e2b-code-interpreter

output

Collecting e2b-code-interpreter
  Downloading e2b_code_interpreter-1.0.0-py3-none-any.whl (15 kB)
Installing collected packages: e2b-code-interpreter
Successfully installed e2b-code-interpreter-1.0.0

Step by step

Use the Sandbox.run_code() method to execute pip install commands inside the sandbox. This example installs requests and verifies the installation by importing it.

python

import os
from e2b_code_interpreter import Sandbox

# Initialize sandbox client with API key from environment
sandbox = Sandbox(api_key=os.environ["E2B_API_KEY"])

# Install 'requests' package inside sandbox
install_code = """
import subprocess
subprocess.run(['pip', 'install', 'requests'], check=True)
"""
install_result = sandbox.run_code(install_code)
print("Install output:", install_result.text)

# Verify installation by importing requests
verify_code = """
import requests
print(requests.__version__)
"""
verify_result = sandbox.run_code(verify_code)
print("Requests version:", verify_result.text)

# Close sandbox session
sandbox.close()

output

Install output: Collecting requests\n  Downloading requests-2.31.0-py3-none-any.whl (62 kB)\nInstalling collected packages: requests\nSuccessfully installed requests-2.31.0
Requests version: 2.31.0

Common variations

Run asynchronous code with await sandbox.run_code() in async functions.
Install multiple packages by chaining pip install commands or listing packages separated by spaces.
Use sandbox.files.write() to upload requirements files and install via pip install -r requirements.txt.

python

import asyncio
import os
from e2b_code_interpreter import Sandbox

async def async_install():
    sandbox = Sandbox(api_key=os.environ["E2B_API_KEY"])
    code = """
import subprocess
subprocess.run(['pip', 'install', 'numpy', 'pandas'], check=True)
"""
    result = await sandbox.run_code(code)
    print("Async install output:", result.text)
    await sandbox.close()

asyncio.run(async_install())

output

Async install output: Collecting numpy\n  Downloading numpy-1.26.4-cp38-cp38-manylinux_2_31_x86_64.whl (17.3 MB)\nCollecting pandas\n  Downloading pandas-2.1.1-cp38-cp38-manylinux_2_31_x86_64.whl (12.5 MB)\nInstalling collected packages: numpy, pandas\nSuccessfully installed numpy-1.26.4 pandas-2.1.1

Troubleshooting

If package installation fails, check the install_result.text for error messages such as missing system dependencies.
Ensure your sandbox session is active; call sandbox.close() after finishing to free resources.
Use subprocess.run(..., check=True) to raise exceptions on install errors for easier debugging.

✅

Key Takeaways

Use Sandbox.run_code() to run pip install commands inside the E2B sandbox securely.
You can install multiple packages or use requirements files by uploading them with sandbox.files.write().
Always close the sandbox session with sandbox.close() to release resources.
Check the output text for errors if installation fails and use check=True in subprocess calls for error detection.

Verified 2026-04

Verify ↗