GPU compute price arbitrage — ONE $0.25 call returns live $/GPU-hour for one GPU class (H100, H200, B200, A100, L40S, RTX 4090/5090) across the vast.ai marketplace floor, RunPod co

GPU compute price arbitrage — ONE $0.25 call returns live $/GPU-hour for one GPU class (H100, H200, B200, A100, L40S, RTX 4090/5090) across the vast.ai marketplace floor, RunPod community/secure cloud, and the AWS on-demand anchor, plus the spread multiple and 24/7 monthly savings versus the anchor. Marketplace-vs-hyperscaler spreads of 3-4x are routine in 2026 — for agents provisioning training or inference this is direct spend recovery. Live provider APIs, no LLM in the path.

price-ascending: candidates are ordered by the listed price of the operation, with no managed-first preference applied. Unlike /v1/apis, this surface does not rank Apiosk-settled listings above federated ones — a comparison that reorders on commercial grounds is not a comparison. Pass sort=managed_first for the catalog's ordering.

Compared: listed price, settlement rail and input compatibility. Not measured yet: latency, reliability, result quality, and provider terms such as rate limits, jurisdiction and licensing.