A project that allocates 50% of its tokens to nodes—what will ultimately sustain these rewards? This was the question I was most curious about while researching @dgrid_ai's tokenomics. In many Web3 projects I’ve looked into, directly allocating half of the token supply to incentivize nodes is pretty uncommon. According to DGrid's latest tokenomics release, the total supply of DGAI is capped at 1 billion tokens: - 50% for nodes, with a 10-year release schedule and halving emissions every two years; - 15% for the community, locked for 6 months, then linearly released over 2 years. The majority of the allocation is reserved for nodes and the community. This structure is clearly designed for a long-term AI service network, not just for a one-time TGE. But a high allocation for nodes doesn’t automatically mean it’s beneficial. Once these node rewards are released, who will continue to need DGAI? DGrid’s proposed closed-loop system is as follows: Users call LLMs or Agents → Pay for AI inference and Agent services using DGAI → Nodes and AI service providers stake DGAI to take on tasks → The protocol distributes rewards based on processing volume, latency, and uptime → Malicious nodes are penalized with a 5%-20% slash of their stake, and the slashed portion is burned. So, DGAI isn’t just an airdrop or governance token—it also serves as a settlement currency for AI services, a staking mechanism for nodes, and a network incentive. DGrid is currently building AI Gateway, AI Arena, DClaw, and a marketplace for models and Agents. Whether these products can generate enough real AI calls and paid usage will determine if the demand side can sustain the long-term token emissions from the node side. If demand for AI usage continues to grow, this “payment-staking-reward” cycle might actually start to work. The project team recently disclosed that DGrid served over 15,000 paying users in the first half of 2026, generating verified revenue of $23 million. This means they’re not starting from zero in creating token demand. However, how much of this business revenue can ultimately translate into sustained demand for DGAI will depend on future metrics like actual payment ratios, node earnings, and token emission data. In my opinion, the key metrics to track for DGAI are: Real usage volume, payment scale, node revenue, and the actual token release rate. The official DGAI airdrop claim portal is now live. You can check your eligibility, view your allocation, choose a claim option, and link your supported exchange account. Claim deadline: August 22, 2026, 08:00 UTC Official claim portal: https://dgrid.ai/airdrop Latest tokenomics: https://(x.com)/dgrid_ai/status/2089601239374069782 In the current market environment, this is a relatively rare project worth keeping an eye on. At least in terms of tokenomics design, DGrid is attempting to support external demand with real AI service revenue, rather than relying solely on staking and token emissions to create an internal loop.