io.net Unveils Incentive Dynamic Engine to Revamp Tokenomics

io.net, a decentralized physical infrastructure network (DePIN) focused on AI compute, has announced a significant overhaul of its tokenomics with the introduction of the Incentive Dynamic Engine (IDE). This new demand-driven model aims to provide long-term stability for suppliers, users, and token holders by moving away from fixed token emissions. Instead, it will implement an automated system that adjusts emissions, buys back tokens, and burns a portion of revenue to mitigate inflationary pressures. Since its launch in June last year, io.net has processed over $20 million in verifiable compute leases, demonstrating real usage demand and necessitating this transition from a bootstrapping phase to a more stable foundation for enterprise and research workloads.
At the core of IDE is a shift from a supply-driven rewards schedule to a real-time, demand-driven control system. This mechanism utilizes two counter-cyclical vaults and a “sustainability ratio” to balance payouts and reserves automatically. When revenue is high, the system retains tokens in reserve, while in times of lower demand, it releases tokens to stabilize supplier payouts. This approach aims to make supplier income predictable and enhance the network’s resilience across varying market cycles, addressing the risks associated with the previous fixed-emissions model.
Moreover, IDE incorporates a built-in deflationary mechanism, committing to using at least 50% of remaining revenue to purchase and permanently burn $IO tokens, with a target of removing over 150 million tokens from circulation over time. The proposal is currently open for community review, with feedback being collected until late February, and a final design is expected to be published ahead of a planned rollout in Q2 2026. Gaurav Sharma, CEO of io.net, emphasized the importance of this shift in enabling decentralized compute markets and aligning incentives for GPU providers, users, and investors, ultimately fostering a reliable and open compute network for AI systems.
Related News





