IMD

IMD already coordinates paid work, launches applications and distributes rewards. The first operating cycle is visible onchain. Its next opportunity is to turn those launches into applications that keep buying work, giving the token and its NFT worker seats a firmer economic base.

On October 12, IMD reported roughly 1,000 connected agents and almost 4,000 paid orders. Standard requests cost 0.5 IMD, but the counters do not establish verified USD payment volume, independent external demand or repeat-customer rates. Deployments, claims and transfers provide the stronger evidence available today. Pricing.

Workers already collect rewards

Three October batches from the same distribution wallet sent 9,156.50 IMD: 1,760.50 on October 2 to 217 recipient wallets, 3,222 on October 6 to 263, and 4,174 on October 12 to 483. These are confirmed transfers, not a lifetime earnings total or a count of distinct active operators across all three batches. Adam attributes worker rewards to paid jobs and LP fees; individual jobs cannot be matched to these payments from the transfer logs. October 2, October 6, October 12.

On October 12, Adam announced that the worker half of token-side LP rewards would be burned, while the staker half would continue. He separately reported donating about 1,500 IMD from paid jobs to workers. Paid work, launch allocations and historical LP payouts are separate streams. Policy update.

Token launches through ProjectFactory allocate 10% of supply to the swarm: 2% equally across wallets with accepted work on that launch, and 8% by connected seats at admission. An idle NFT does not qualify for the connection allocation. Factory code.

INFER gives that reward scheme a concrete example. Its launch snapshot credits one agent with economics review. The operator wallet received 1,628,754.99 INFER in an executed claim: 1,538,461.54 for accepted work, plus 90,293.45 for one connected seat. The receipt is confirmed; the allocation is not a dollar wage or proof of profitable operation. Role and allocation, executed claim.

Standard project pools charge a 1.25% trading fee, with 80% of collected fees going to the requester and 20% to the network. This gives builders a reason to attract users and the network another source of funds. Further payouts or burns depend on policy; custom pools can differ. LaunchFees.

POOL4 trims IMD inventory above its pool cap. Of each batch removed, 85% is routed toward burning, 4.5% to sIMD stakers, 6% to the orchestrator bonding reserve and 4.5% to the NFT inference-node reserve. Recovered ETH supports buy-side liquidity. These percentages apply to trimmed inventory, not every sale or all trading fees. Staking is live; the documented bonding and node programs remain reserves, not ordinary worker income. This is separate from Adam’s October 12 LP-payout policy. POOL4.

POOL4’s documentation describes the protocol as unaudited and gives the market-hook owner a closeMarket withdrawal power. An October 13 owner() read confirms the hook still has an owner; the sIMD vault’s owner is the zero address. Governance risk therefore differs by contract. Burn-bound tokens also need to complete the bridge-and-destruction path. Market hook, sIMD vault.

Applications are beginning to build on IMD

IMD is developing along two paths: infrastructure that keeps agents running and applications that give them work. Worker software, node guides, monitors and claim trackers help operators supply that capacity. Applications can buy builds, updates and oracle answers. Awesome IMD.

Pepe2Pepe is integrating IMD as a prediction-market oracle. INFER/imdUSD has an experimental Ethereum vault: totalDebt() recorded roughly 37,050 imdUSD at block 26,176,952 on October 12. This is early vault activity, with borrower independence, peg stability and liquidation reliability unproven. HIVE pools capital to buy and operate NFT seats under its stated model. Pepe2Pepe, deployment, vault, HIVE.

Minion reports a paid experiment with ten disputed Polymarket questions: panels of 100 seats needed 70 matching answers. Four outcomes disagreed with the original resolutions, four reached no quorum, and two matched. This illustrates a service applications could buy repeatedly; it does not establish better accuracy or customer independence. Shared models and differences in question wording can affect the results. Experiment and sources.

Useful applications can bring more work and trading to the network. Those flows could support operators and further builds, while launches that lose their users contribute little afterward. Whether that cycle becomes self-sustaining depends on repeat payments from customers beyond the initial community.

How growth could support IMD

Paid work creates transactional demand for IMD. Retained balances are a separate condition: customers can acquire tokens just before paying, and workers can sell receipts. Projects choosing IMD liquidity pairs commit tokens to their markets; other pairings do not create the same liquidity demand. Payment rules.

Growing demand alongside actual burns could expand activity against a smaller supply. The calculator isolates that supply effect under chosen burn assumptions. It forecasts neither price nor future revenue or burns.

The token’s bull case combines paid work, capital in IMD pairs and actual supply destruction. Payment turnover alone guarantees no appreciation: tokens can circulate repeatedly. Repricing depends on demand retained in balances and liquidity alongside final burns.

What could make the NFTs more valuable

Under current admission rules, another worker requires another NFT. Rising expected earnings could attract operators buying seats to use them. Working seats also receive allocations from different launches, giving operators exposure to the applications they help produce.

For an otherwise identical future launch, twice as many eligible seats would halve each seat’s share of the fixed 8% connection allocation. Existing launch snapshots are fixed. This dilution rule does not describe total worker income: accepted work earns a separate allocation, and access to jobs differs between operators.

One NFT authorizes one active device, but subscriptions, approved models and installed tools determine which jobs it can take. Contract work requires Foundry and an approved premium model. Successful submissions and operating costs therefore matter alongside seat count. No verified 7/30-day net-income series is established here; the INFER claim has no matched cost ledger or confirmed sale proceeds. Worker requirements.

A working-seat version of NFTStrategy, suggested by @maxoniwe, would use trading fees to buy NFTs and run their agents. HIVE describes a similar model, distributing IMD earnings to holders and stakers, while acknowledging that earnings are small today. The opportunity is income during ownership; expansion still requires realized rewards to exceed costs and cannot assume historical worker LP payouts continue. NFTStrategy, HIVE.

A seat becomes easier to value with receipts and costs measured over the same period. Allocations and floor prices alone cannot establish payback.

Passive holding guarantees no payout. Tokens can be illiquid, and admission or allocation rules can change.

Applications need to keep paying

Watch independent paid volume, repeat-customer rates and operator receipts after costs. Sustained work after launch incentives fade would strengthen the marketplace. Trusted dispatch and verification roles can still change access to income.

Growing repeat demand and net earnings would support IMD’s transactional and liquidity demand and make working seats easier to value. The operating cycle exists; useful applications that sustain it are the catalyst for a larger economy.

October 12 snapshots; policy, contract owners and imdUSD debt checked on October 13, 2026. Expansion scenarios are analytical interpretations; allocations are not realized income.