Skip to main content
This quickstart creates one sandboxed computer, runs uname -a, prints the output, and deletes the computer. You can use the language selector to switch the whole example between Python and TypeScript.

Before you start

Get an API key from celesto.ai under Settings > Security. Set it as CELESTO_API_KEY before you run the examples.

Install the Python SDK

Run your first computer

A named API key needs all three permissions for this example: See your computers, Create, start, stop computers and run commands, and Delete computers. The final permission lets the example clean up after itself.
The free trial allows one sandbox at a time, including stopped sandboxes. Reuse your existing sandbox, delete it when you are done, or add prepaid credit before creating another. HTTP 429 can mean this limit; waiting does not free the slot.
Create quickstart.py:
quickstart.py
Run it:
The script prints the computer ID and Linux system information. Exiting the with block deletes the computer, including when your code fails.
If creation fails, follow the error’s recovery message. If the response is unreadable or the connection times out, inspect your computers in the dashboard before creating another. See error handling for uncertain outcomes and cleanup failures.

Next step

Learn how to create computers with templates when your agent needs coding tools already installed.
Last modified on October 22, 2026