ScoutWyze Compute

Rank current RunPod GPU offers for your workload. Timestamped quotes and scores. $10 prepaid, no subscription.

1. Try it free — no key required

curl https://scoutwyze-compute.fly.dev/v1/compute/sample

2. Real query with your own key

curl -X POST https://scoutwyze-compute.fly.dev/v1/compute/rank \
  -H "Authorization: Bearer sw_live_..." \
  -H "Content-Type: application/json" \
  -d '{"gpuClass":"H100","preference":"cheapest"}'

Example response

{
  "status": "ok",
  "schema_version": "1.0",
  "coverage": { "vertical": "gpu_compute", "providers_live": ["runpod"] },
  "recommended": {
    "provider": "runpod",
    "sku": "NVIDIA H100 80GB HBM3",
    "region": "US-CA-2",
    "vendorHourly": 21.2,
    "observed_at": "2026-09-23T00:17:35.911Z",
    "freshness_seconds": 41,
    "source": "live_api",
    "availability_status": "low",
    "classification": "provider_reported",
    "score": 0.999,
    "reason": "Cheapest match on runpod: $21.20/hr (price score 1.00), updated just now."
  },
  "alternatives": [ ... ],
  "limits": { "not_reserved": true, "not_provisioned": true, "can_provision": false },
  "billing": { "billable": true, "unit": "successful_rank", "price_usd": 0.15, "rail": "bearer", "creditsRemaining": 9.85 }
}

RunPod offers below are sourced from RunPod's own live catalog API, refreshed continuously. Lambda Labs/CoreWeave rows (visible via the source field) are fixture data for comparison only — never live, never bookable.

POST /v1/compute/rank — $0.15/request. Dual-rail: Bearer key (debited only on a real match) or x402/USDC on Base (settles on payment, before scoring — see billing.rail in the response). llms.txt has the full asymmetry explained.

Your API key (shown once — save it now):

This cannot be shown again. Store it before continuing.

Pick a credit pack to fund your account:

Questions or something broken? support@scoutwyze.com · llms.txt · openapi.json