AI token spending jumps 122% to 160% as prices fall 67%, a16z data shows

AI token spending surged 122% to 160% even as token prices fell 67%, according to a16z data, signaling accelerating AI adoption across crypto networks despite collapsing unit economics. The divergence suggests usage is growing far faster than revenue capture, with falling prices likely reflecting intense competition among inference providers and protocols. Rising GPU costs compound the margin pressure, potentially challenging profitability for AI infrastructure projects that depend on compute subsidies. Key protocols in this sector include Bittensor (TAO), Render (RNDR), Akash (AKT), and Near (NEAR), which operate across decentralized compute and inference markets. The trend matters because it tests whether decentralized AI networks can sustain growth without continued token emissions or venture funding. Watch next for GPU pricing trends, further token price declines, and whether spending growth can outpace margin compression to restore sustainable economics.