ERA Cloud

ERA Cloud API quickstart.

A small operational guide for approved private-beta tenants and operators: provision access, sync providers, create GPU, VPS/CPU or AI workloads, compare market prices, and estimate spend.

The reliable 1-2-3 path

The product should feel simple even while provider routing, billing and reconciliation stay rigorous behind the scenes.

PATH1. Get approved access

ERA issues a tenant, API key, dashboard URL, base URL and first smoke command. No public signup maze.

PATH2. Pick the job type

Use one API for AI routing, GPU jobs, VPS/CPU, dedicated compute, storage or network resources.

PATH3. Let ERA route and reconcile

ERA selects the cheapest viable provider, records usage, estimates charges and reconciles provider statements.

1. Provision tenant

Approved beta operators create the tenant and capture the returned JWT and API key.

curl -X POST https://api.eracloud.pro/api/v1/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "tenant_name": "Acme AI Lab",
    "email": "ops@acme.ai",
    "password": "change-me-strong"
  }'

2. Create routed workload

Ask ERA Cloud to pick the cheapest viable resource provider in the requested region.

curl -X POST https://api.eracloud.pro/api/v1/workloads \
  -H "Authorization: Bearer $ERA_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "tenant_id": "ten_...",
    "kind": "server",
    "profile": "gpu-h100",
    "region": "us-east",
    "routing_policy": "cheapest"
  }'

3. Connect coding tools

ERA Cloud exposes an OpenAI-compatible endpoint for IDEs and AI coding agents. Set the base URL once, paste an ERA API key, and let ERA route each request to the cheapest viable provider.

{
  "provider": "ERA Cloud",
  "baseUrl": "https://api.eracloud.pro/v1",
  "apiKey": "$ERA_API_KEY",
  "model": "deepseek-v4-pro"
}

4. Route AI work from the IDE

Use provider-agnostic model names and optional preferred providers; ERA returns routing metadata with each response.

curl https://api.eracloud.pro/v1/chat/completions \
  -H "Authorization: Bearer $ERA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v4-pro",
    "messages": [
      { "role": "user", "content": "Refactor this function safely." }
    ],
    "preferred_providers": ["fireworks", "deepinfra"]
  }'

5. Discover remote operations

API clients and future MCP wrappers can fetch one manifest that explains available tools, REST paths, risk level, idempotency, schemas and safety notes.

curl https://api.eracloud.pro/api/v1/remote/operations \
  -H "Authorization: Bearer $ERA_API_KEY"

6. Manage provider instances safely

Instance list/create/stop APIs are available for live providers. Unsupported or unsafe actions return clear errors instead of fake success.

curl https://api.eracloud.pro/api/v1/providers/runpod/instances \
  -H "Authorization: Bearer $ERA_API_KEY"

Developer tool coverage

These connector profiles are treated as first-class product surfaces, not afterthought integrations.

READYCodex and OpenAI-compatible coding agents

Use https://api.eracloud.pro/v1 as the OpenAI-compatible base URL and an ERA Cloud API key as the bearer token.

READYCursor custom provider settings

Use https://api.eracloud.pro/v1 as the OpenAI-compatible base URL and an ERA Cloud API key as the bearer token.

READYVS Code / VSCodium extensions such as Continue, Cline and Roo Code

Use https://api.eracloud.pro/v1 as the OpenAI-compatible base URL and an ERA Cloud API key as the bearer token.

READYOpenCode and other CLI-first coding agents

Use https://api.eracloud.pro/v1 as the OpenAI-compatible base URL and an ERA Cloud API key as the bearer token.

READYGeneric OpenAI-compatible clients

Use https://api.eracloud.pro/v1 as the OpenAI-compatible base URL and an ERA Cloud API key as the bearer token.

Core endpoints

These are the endpoints to smoke-test immediately after the Render API deploy.

GET/health

Render and uptime checks. Public, outside auth and rate limiting.

POST/api/v1/auth/register

Operator-only beta provisioning for approved tenants, users, JWTs, and first API keys.

POST/api/v1/providers/{provider}/sync

Load provider capabilities into the control plane.

POST/api/v1/workloads

Create a routed server or inference workload.

GET/api/v1/developer-tools/connectors

Return ready-to-use connector profiles for Codex, Cursor, VS Code/Codium, OpenCode and OpenAI-compatible agents.

GET/api/v1/benchmark/gpu

Compare normalized GPU prices across known provider capabilities.

GET/api/v1/billing/estimate

Estimate daily and monthly spend for a tenant.

GET/api/v1/remote/operations

Discover REST operations, MCP tool names, risk levels, idempotency, input schemas, and safety notes.

GET/api/v1/mcp/manifest

MCP-ready manifest for remote clients that need tool names and safe instance-management contracts.