Guide
API
Blog
Version
Guide
API
Blog

Jupyter

AIO Sandbox includes Jupyter kernel execution for Python code. It is useful for data processing, charts, notebooks, and stateful Python workflows.

Execute Code

curl -X POST "http://localhost:8080/v1/jupyter/execute" \
  -H "Content-Type: application/json" \
  -d '{
    "code": "x = 40 + 2\nprint(x)"
  }'

Stateful Sessions

Use session_id to keep variables across requests:

curl -X POST "http://localhost:8080/v1/jupyter/execute" \
  -H "Content-Type: application/json" \
  -d '{
    "session_id": "analysis",
    "code": "data = [1, 2, 3]"
  }'

curl -X POST "http://localhost:8080/v1/jupyter/execute" \
  -H "Content-Type: application/json" \
  -d '{
    "session_id": "analysis",
    "code": "print(sum(data))"
  }'

Kernel Info

curl "http://localhost:8080/v1/jupyter/info"

Session Management

curl "http://localhost:8080/v1/jupyter/sessions"
curl -X DELETE "http://localhost:8080/v1/jupyter/sessions/analysis"

Tips

  • Use Jupyter for Python code that benefits from stateful variables or rich output.
  • Use /v1/code/execute when you want a language-neutral entry point.
  • Store generated files in the workspace so browser, shell, and file APIs can access them.