Developer Platform & SDKs

Developer Portal & API Keys

Integrate distributed compute nodes directly into your training pipelines with official SDKs and CLI.

Active API Keys

Scoped Bearer authentication for REST & SDK

Key Name	Token Prefix	Permissions	Created	Last Used	Actions
Default CI/CD Training Key	dc_live_7fa9...1238	WORKLOADS_ALLNODES_READBILLING_READ	8/19/2026	3:57:57 PM

SDK Quickstart & Code Recipes

Select your preferred language or toolchain to get started in seconds.

import { DistributedComputeClient } from '@distributed-compute/sdk';

const client = new DistributedComputeClient({
  apiKey: process.env.DISTRIBUTED_COMPUTE_API_KEY,
});

// 1. Discover top-tier GPUs
const { nodes } = await client.nodes.list({
  tier: 'TIER_1_ENTERPRISE_GPU',
  minVramGb: 48,
});

// 2. Submit a high-performance training container
const job = await client.workloads.submit({
  image: 'pytorch/pytorch:2.1.0-cuda12.1-cudnn8-runtime',
  command: 'python train_distributed.py --epochs 50',
  gpuCount: 4,
  strategy: 'BEST_PRICE_PERFORMANCE',
});

console.log(`Job submitted: ${job.id} on node ${job.nodeId}`);