GPU.ai as cloud orchestrator - Per-second billing, zero markup claim, 12+ cloud aggregation

Fetched 2026-10-04. GPU.ai is a cloud aggregator/orchestrator that reads live pricing from 12+ GPU clouds and normalizes to per-GPU-hour rates. Key structural facts: (1) Per-second billing - no minimum hour commitment. (2) Zero markup claim - rates are raw provider rates. (3) Two tiers: SECURE (vetted datacenter capacity) and COMMUNITY (independent marketplace suppliers). (4) Stock counts are launchable inventory, not theoretical capacity. (5) Data refreshes hourly via public no-auth JSON API. (6) 1,692 GPUs launchable across 27 models as of 2026-10-04. (7) The 'Market range' column shows the span of public on-demand list prices across major clouds for the closest equivalent SKU. GPU.ai is the most transparent aggregator in the market - it shows both its own rates AND the market range, making it possible to see the spread between aggregator and raw cloud rates. The main limitation is that GPU.ai is itself a provider, so its 'secure' rates may include some markup despite the 'zero markup' claim.