The Grid Wants to Auction Your LLM Requests

Field notes

The Grid describes inference as a live clearing market rather than a fixed model-selection or subscription workflow. Its stated sequence is that suppliers bid for requests, users select an inference tier, and the resulting charge is calculated per token at the market price. The homepage claims savings of up to 80% and a path from zero to a first request in under five minutes, but does not identify suppliers or expose market activity.