Run LLMs on the world's spare GPUs

An OpenAI-compatible inference marketplace — pay per token at live market prices, or sell your own GPU capacity.