SHIP AGENTS
THAT EXECUTE
CODE.
Your agents need a computer and a drive. Here's the API.
USE CASES(you can build in a weekend!)
Code Interpreter
Give users a personal data analyst. Agent writes Python, sandbox runs it securely.
TRUSTED RUNTIME
FOR UNTRUSTED CODE.
One API for code, sessions, and artifacts.
Sandboxed by default.
Run real code
Allow your AI agents to run Python and shell commands inside a safe workspace.
Install what your agent needs
If your AI needs libraries or system tools, just list them. They’re installed automatically and reproducibly for that run.
Files in. Files out.
Upload data for your agent to process. Download everything it creates—charts, reports, transformed datasets.
Plug into any agent
Use it as a single “tool” inside any agent framework. Works client-side or server-side with minimal setup.
Start building.
Initialize your environment in seconds.
Ready for deployment.
Read https://agentsandbox.co/skill.md and follow the instructions to start using Agent Sandbox
Simple, Transparent Pricing.
Pay only for what you use. No subscriptions, no commitments.
$10 free credits when you sign in with Google(~11 hours of compute)
Compute
Billed by the second
Storage
For output files
How it works
Add credits
Purchase anywhere from $5 to $1,000 at a time
Run your code
Credits are deducted automatically based on actual usage
Track everything
See your balance, usage history, and spending patterns in your dashboard