IO has a hard cap of 800 million tokens, no minting authority to override it, and since June 2026 a burn engine that permanently destroys at least half of the network's post-payout revenue. That combination makes io.net's token one of the few DePIN assets whose supply can actually shrink relative to plan when business is good, which is the detail most tokenomics summaries miss.
The other half of the story is less flattering: 300 million of those tokens are still being emitted to GPU suppliers and stakers on a 20-year schedule, so IO's net supply picture is a race between programmed inflation and revenue-linked burns.
The 800 Million Cap and What Circulates
The token generation event in June 2024 created 500 million IO, with the remaining 300 million reserved for hourly block rewards paid to hardware suppliers and their stakers. As of July 2026, circulating supply sits near 362 million, about 45 percent of the maximum, per the project's published figures and market trackers.
Circulating supply grows from two directions at once. Emissions drip out hourly to the people keeping GPUs online, and the allocations from launch continue vesting to investors, contributors, and ecosystem programs. Anyone modeling IO's float needs both curves, since either one alone understates future sell-side supply.
Who Got What at Launch
The official allocation breakdown splits the 800 million cap as follows:
| Category | Share of supply | Notes |
|---|---|---|
| Supplier and staker emissions | 37.50% | 300M paid hourly over ~20 years |
| R&D and ecosystem | 16.01% | Development, grants, operations |
| Seed investors | 12.50% | Earliest backers, vesting |
| Initial core contributors | 11.34% | Team allocation, vesting |
| Series A investors | 10.15% | 2024 round at a $1B valuation |
| Community programs | 10.00% | Airdrops and incentive seasons |
| Launch campaign rewards | 2.50% | Distribution at listing |
Investor and team tranches follow multi-year vesting, and their unlock events have historically coincided with IO's sharpest drawdowns. That pressure fades with time as the vested share approaches completion, which is one reason the supply outlook improves the further you get from the 2024 launch.
Emissions: Hourly Rewards on a 20-Year Clock
io.net pays block rewards every hour to suppliers who keep verified hardware online, drawn from the 300 million emissions pool. The schedule is disinflationary: it started at 8 percent annual inflation in year one and steps down roughly 1 percent each month until the cap is reached in the 2040s. Rewards scale with GPU model, uptime, and stake, and every device requires its supplier to lock at least 200 IO, which recycles part of the emitted supply straight back into staking collateral.
This design borrows from Bitcoin's fixed-cap discipline while paying for something other than hash power. The network buys real capacity with predictable inflation, and the question for holders is simply whether each emitted token generates enough revenue to justify its existence.
The Incentive Dynamic Engine Changes the Math
The June 2026 overhaul, called the Incentive Dynamic Engine, made two structural changes. Emissions now adjust dynamically to network demand instead of following the static curve alone, and at least 50 percent of post-payout network revenue in IO is burned permanently, with io.net projecting up to 12 million tokens destroyed in the engine's first year. CoinDesk's research note on the engine frames it as the network tying token supply to commercial reality for the first time.
The contrast with peers is instructive. Render's burn-and-mint model burns payment tokens but mints rewards against them, while IO's burns subtract from a capped supply, so sustained usage compounds scarcity rather than recycling it. A Messari report on the new tokenomics estimates that at healthy sustainability ratios, burns can offset half or more of ongoing block-reward emissions.
What the Flows Mean for Holders
Put the pieces together and IO's supply math has three regimes. In a low-revenue regime, emissions dominate and the float grows about as fast as the schedule dictates. In a balanced regime, burns absorb a meaningful share of new supply and the float grows slowly. In a high-revenue regime, burns plus staking demand exceed emissions and the effective float contracts.
Which regime prevails is observable on-chain quarter by quarter, and it feeds directly into IO's price outlook, where the gap between bearish and bullish forecasts largely reduces to this one variable.
Frequently Asked Questions
Is IO inflationary or deflationary?
Both, depending on network revenue. The emissions schedule adds new IO hourly through the 2040s, while the Incentive Dynamic Engine burns at least half of post-payout revenue, so the net direction flips deflationary whenever burns and staking demand outpace the declining emission rate.
How many IO tokens are locked in staking?
Every active device on the network requires a minimum stake of 200 IO from its supplier, adjusted upward for higher-end GPU models, and co-stakers add more on top. The total moves with hardware onboarding, so the staked share of supply rises as the verified GPU count grows.
When will all 800 million IO be in circulation?
The emissions pool pays out over roughly 20 years from the June 2024 launch, putting full circulation in the mid-2040s. Investor, contributor, and ecosystem allocations vest much sooner, so the vast majority of non-emission supply unlocks during the 2020s.
Supply Discipline as io.net's Second Product
IO tokenomics started as a fairly standard DePIN design, a capped supply funding hardware incentives, and matured in 2026 into something stricter: demand-adjusted emissions with a contractual burn tied to revenue. The token now behaves like a claim on network activity rather than a pure incentive chip.
For traders, the practical takeaway is to watch the sustainability ratio and cumulative burn figures the project publishes, because they convert directly into net supply growth.
You can act on that view by buying IO on spot or trading IO perpetual futures on LeveX, and find breakdowns of other token models in Crypto in a Minute.