What this desk shows

Public list prices published by an aggregator feed: GPU hourly rates from RunPod, Vast.ai, AWS, Azure and the neoclouds in the Shadeform catalogue, and LLM token prices from five inference routers. An observation display: not an executable offer, not a quote, and nothing here can buy, reserve, or settle anything.

What it does not claim

Provenance

Data: GPUDiff public GPU offer feed, CC BY 4.0. Imported, hashed, and normalized by the HyperAI data layer. Local replay-level evidence only. Neocloud rows and capacity come from the Shadeform public catalogue, shown on the operator's confirmation that display is permitted (terms); availability is point-in-time, not a reservation. Layout inspired by Desk.