v1 REST API

Getting Started

Welcome to the Nebula Compute developer API reference. Our REST interface enables developers to remotely provision enterprise-grade GPU enclaves on-demand in under 90 seconds. This getting-started guide will walk you through authenticating your requests and orchestrating high-performance AI workloads directly on our bare-metal hardware clusters in Santo Domingo, Dominican Republic.

💡

Quick Start Details

Generate your secure API private key from the Nebula console dashboard at nbl.do, then construct all HTTPS calls containing the Bearer token in the 'Authorization' header.

Authentication

Every API call directed to Nebula's endpoints requires secure authorization. Transmit your private credential within the dynamic HTTPS headers as a Bearer token.

                        Terminal (cURL)
                    
                        curl -X GET "https://nbl.do/api/v1/instances" \
  -H "Authorization: Bearer nbl_sec_key_YOUR_TOKEN"

⚠️

Industrial Grade Security

Never commit API access tokens to raw public repositories. Always leverage environment variables to dynamically inject authorization credentials.

Base URL Endpoint

The global, high-performance REST routing gateway for all Nebula programmatical structures is located at:

URL https://nbl.do/api/v1

Locations & Hostnodes Discovery

Retrieve dynamic listings of active physical hostnodes, available regions, and hardware cluster availability in our Tier-III DR facility.

GET /locations

                        Response Object JSON
                    

                        {
  "ok": true,
  "regions": [
    {
      "id": "sd-tier3-1",
      "name": "Santo Domingo Tier-III Node",
      "available_gpus": {
        "NVIDIA-H100-PCIE-80GB": 12,
        "NVIDIA-A100-SXM-80GB": 8,
        "NVIDIA-RTX-A4000-16GB": 32
      }
    }
  ]
}
                    

Instance Creation

Request and deploy fully isolated virtual enclaves or dedicated GPU systems equipped with high-speed SSD arrays and dynamic root shell credentials.

POST /instances/provision

                        Payload Parameters JSON
                    

                        {
  "gpu_type": "NVIDIA-H100-PCIE-80GB",
  "gpu_count": 1,
  "operating_system": "Ubuntu-22.04-CUDA-12.1",
  "storage_gb": 500
}
                    

Instance Management

Operate instances programmatically. You can stop computation cycles, query telemetry stats, or release enclaves (performing dynamic SSD sanitization).

POST /instances/{id}/power

                        Payload Parameters JSON
                    
                        {
  "action": "stop" // Options: start, stop, restart, pause
}

Rate Limiting

To preserve latency and network isolation thresholds across the Caucedo Gateway, requests are bounded to a maximum of 120 calls per minute per organization.

SLA & Dedicated Support

All provisioned API gateways and virtual hardware nodes remain covered under our robust SLA guaranteeing 99.99% network availability. On-site engineers remain available 24/7/365 in our Santo Domingo Tier-III operations facility.