NVIDIA Air SDK#
A Python SDK for creating, running, and managing network simulations on NVIDIA Air.
Installation#
The package is available on PyPI.
pip install nv-air-sdk
Or with uv:
uv add nv-air-sdk
Requires Python 3.10+.
Authentication#
The SDK supports three authentication methods. See the NVIDIA Air authentication guide for details on obtaining credentials.
API Key — pass an NGC Scoped API Key (SAK) directly:
from air_sdk import AirApi
api = AirApi.with_api_key(api_key='nvapi-xxxx')
NGC Config — reads the key from ~/.ngc/config, created by
ngc config set (requires the NGC CLI):
api = AirApi.with_ngc_config()
# or equivalently:
api = AirApi()
Device Login — opens a browser for interactive NGC authentication (no pre-existing key required):
api = AirApi.with_device_login(
email='user@nvidia.com',
org_num='my-org',
)
Non-Standard Hostnames#
All factory methods accept an api_url parameter that defaults to the
public NVIDIA Air instance. If you are connecting to a different Air
deployment, browse to /api on the host
(e.g. https://your-air-host.example.com/api) and pass the API URL
from the servers dropdown to the SDK:
api = AirApi.with_api_key(
api_key='nvapi-xxxx',
api_url='https://your-air-host.example.com',
)
Quick Start#
from air_sdk import AirApi
api = AirApi.with_api_key(api_key='nvapi-xxxx')
# List simulations
for sim in api.simulations.list():
print(sim.name, sim.state)
# Create a simulation
sim = api.simulations.create(name='my-sim')