GPU Grid is a decentralized GPU compute network that connects idle graphics processing units with developers who require AI inference capacity. The protocol operates as a peer-to-peer marketplace where GPU owners contribute their hardware to the network and earn $GGRID tokens in exchange for processing compute jobs submitted by developers and applications.

The network routes inference requests across available nodes based on four signals: price, speed, reliability, and current load. Each node maintains a reputation score that determines its priority in job routing. Nodes that fail to process jobs reliably are temporarily removed from the routing pool until their score recovers.

$GGRID is the native token of the GPU Grid network. It is used for compensating GPU providers, staking nodes, governance participation, and buyback-and-burn mechanics. Every completed compute job automatically splits the fee: 75% goes to the GPU provider, 12.5% is allocated to burn, 7.5% to stakers, and 5% to the treasury. The token was launched on Solana via pump.fun and the on-chain payout program has been independently verified.

Developers access the network through an OpenAI-compatible API endpoint, allowing existing applications to route inference to the decentralized network without code changes. GPU providers join the network by running a one-line installer that configures Ollama, opens a secure tunnel, and registers the node with the network.