GRID Docs | General Robotics

Python client for GRID Cortex — cloud model inference for robotics: depth, detection, segmentation, grasping, VLMs, and VLA policies behind one run() call.

Install

# Preinstalled in every GRID session workspace and in the Python environment the
# GRID CLI prepares when you run a program with `skill run`. Nothing to install.

Quickstart

from grid_cortex_client import CortexClient, ModelType

client = CortexClient()  # reads GRID_CORTEX_API_KEY from the environment
depth = client.run(ModelType.ZOEDEPTH, image_input="scene.jpg")

Every hosted model and its inputs/outputs is documented in the Cortex Models section of the sidebar. New client releases are available through GRID session workspaces and the Python environment managed by the GRID CLI. If you previously installed the client from public PyPI, use one of these environments for current versions. Existing public PyPI releases remain available as historical versions; no new releases are published there.

Classes

CortexClient
AsyncCortexClient
CortexExecutionEvent
RuntimeObserver
CortexHubClient
CortexHubError
HubResult

Python API

grid-cortex-client

Install

Quickstart

Classes

​Install

​Quickstart

​Classes

Install

Quickstart

Classes