Public API

Live rates, as JSON

Every number on the homepage comes from these endpoints. They are public, read-only, CORS-enabled and cached for 30 seconds. No API key needed. Prices are USD per GPU per hour. Each response carries as_of (observation time of the market data, null while none has been recorded), stale and stale_sources (which of market or capacity is past its freshness budget). Fields may be null where the market has no value, including max_price_usd_hr; floor_price_usd_hr is the minimum listing price and reference_price_usd_hr the market reference price for the model. Responses carry a weak ETag; send If-None-Match to get a 304 when nothing changed.

GET/api/v1/rates

GPUs we're accepting now: idle pay per GPU-hour and open 1x / 8x spots per GPU tier.

Example values

{
  "ok": true,
  "totals": {
    "open_spots": 117,
    "open_gpu_slots": 607
  },
  "data": [
    {
      "id": "b300",
      "gpu": "B300",
      "reward_usd_hr_per_gpu": 6.4,
      "spots": {
        "1x": {
          "open": 1,
          "total": 4,
          "full": false
        },
        "8x": {
          "open": 1,
          "total": 4,
          "full": false
        }
      },
      "open_gpu_slots": 9,
      "updated_at": "2026-09-25T10:00:00.000Z"
    }
  ]
}

GET/api/v1/market?listed=true&segment=datacenter

GPU market overview for all supported models. Optional filters: listed=true, segment=datacenter|workstation|consumer.

Example values

{
  "ok": true,
  "as_of": "2026-09-25T17:16:45.863Z",
  "stale": false,
  "stale_sources": [],
  "summary": {
    "supported_models": 88,
    "listed_now": 23
  },
  "data": [
    {
      "id": "h100-80gb-hbm3",
      "name": "NVIDIA H100 80GB HBM3",
      "listed": 204,
      "rented_now": 149,
      "utilization_pct": 61.48,
      "utilization_window": "trailing",
      "utilization_now_pct": 73.04,
      "utilization_30d_pct": 61.48,
      "lowest_price_usd_hr": 1.3,
      "highest_price_usd_hr": 2.75,
      "reference_price_usd_hr": 1.39,
      "floor_price_usd_hr": 0.695,
      "price_limit_usd_hr": 3.76,
      "max_price_usd_hr": 5.56,
      "rentable_now": 54,
      "rentable_nodes": 19,
      "idle_incentive": true,
      "idle_incentive_group": "h100"
    }
  ]
}

GET/api/v1/gpus

Supported GPU catalog with idle pay, floor and reference price per model.

Example values

{
  "ok": true,
  "as_of": "…",
  "stale": false,
  "stale_sources": [],
  "data": [
    {
      "id": "b200",
      "name": "NVIDIA B200",
      "vram_gb": 180,
      "idle_reward_usd_hr_per_gpu": 4.25,
      "floor_price_usd_hr": 2.8,
      "reference_price_usd_hr": 5.6,
      "lowest_price_usd_hr": 5.5,
      "rentable_now": 6
    }
  ]
}

GET/api/v1/stats

Headline numbers for dashboards and widgets.

Example values

{
  "ok": true,
  "as_of": "…",
  "stale": false,
  "stale_sources": [],
  "data": {
    "supported_models": 88,
    "models_listed_now": 23,
    "gpus_listed": 1484,
    "gpus_rented_now": 1185,
    "open_spots": 133,
    "open_gpu_slots": 672,
    "top_idle_reward_usd_hr": 6.4,
    "updated_at": "…"
  }
}

POST/api/v1/waitlist

Join the rental waitlist. Body: { email, name?, gpu_interest?, use_case? }.

Example values

curl -X POST https://rackloom.in/api/v1/waitlist \
  -H 'Content-Type: application/json' \
  -d '{"email":"you@company.com","gpu_interest":"H100"}'

Rate limit: 10 waitlist requests per 10 minutes per IP. Need webhooks or higher limits? Email support@rackloom.in.