Build on the network.

Submit distributed inference, build, and agent workloads through the API, SDKs, or command line.

API v1

Install via pip:

pip install planetnodes

Python Integration Code:

from planetnodes import PlanetNodes

pn = PlanetNodes(api_key="pn_live_...")

# Run AI inference on global compute mesh
result = pn.run_inference(
    prompt="Synthesize research data",
    region="us-east",
    verify=True
)

print("Job Result:", result["result"])

# Operations for production integrations
usage = pn.get_usage()
pn.configure_webhook("https://example.com/webhooks/planetnodes")
retry = pn.retry_job("failed_job_id")

38 global sub-regions

Target specific geographic regions (`us-east`, `eu-west`, `ap-southeast`) for low-latency edge compute execution.

2-node verifiable pair

Pass "verify": true to run deterministic jobs on independent provider nodes and compare normalized results.

Multi-stage workflows

Set `parent_job_id` to automatically chain multi-step job pipelines with automatic parent result injection.