NOX is a Telegram-native protocol that rewards people for the effort they already spend, funded only by revenue that has already cleared and never by minting tokens. This paper explains the mechanism, the token, and the design. The full formal treatment lives in the companion NOX Technical Specification.
Technology reaches a person in one of two ways. It is bought, which excludes anyone for whom the price is the binding constraint. Or it is nominally free and funded by harvesting that person's attention and data for sale to advertisers, an arrangement in which the participant is the product and the only return is more precisely targeted advertising. Both prices the same access, and both settle value in the same direction. Under the first, the member who can pay is admitted and keeps the value of admission. Under the second, the member who cannot pay is admitted, but the value their attention generates accrues to the platform rather than to the member who generated it.
NOX is an economic protocol that requires neither. It substitutes a third admission price, effort, and returns the value that effort creates to the member who created it. Effort here is wider than time and attention. It covers attention given to content, tasks and lessons completed, activity and rest recorded from a device the member controls, work the ecosystem incorporates, and data explicitly opted in at the member's discretion. Data is never taken as the uncompensated cost of "free". When a member contributes, the protocol issues a redeemable claim in return, denominated in points and later bridged to an on-chain token, $NOX, that the member directs across a catalog of products and services.
The mechanism that connects contributed effort to the return it earns is one this document names Proof of Clearing: reward issued in proportion to a member's verified contribution, but only against revenue that has already settled. The condition the name records is load-bearing. Contribution-reward schemes are not new, but they characteristically fund issuance from a token minted for the purpose. Here it is the clearing of revenue that licenses issuance, so a claim can be created only once the revenue backing it has arrived, whether from demand-side partners or from member subscriptions, and never from token emission. Proof of Clearing generalizes the reward-for-demonstrated-work principle of Proof of Work while inverting its admission condition: no capital, hardware, or energy is required to participate, only contribution itself.
Reward issuance rests on three ways of participating. The first is access and attention, engagement across content and offers that demand-side partners pay to reach. The second is productive contribution, work the ecosystem incorporates, rewarded in proportion to the value delivered. The third is routed demand, a purchase the member intended anyway, directed through the protocol and rewarded from the commission it generates. A member may also elect to pay a subscription for ad-free access, which contributes cleared revenue that adds to the reward pool rather than competing with it.
Several properties follow from funding reward in arrears, and they are stated as claims here and proved formally in the NOX Technical Specification. Rewards are funded only by cleared revenue and never by emission. The reward pool cannot go negative, because it is filled before it is drawn. Weighting contributions changes how the pool is divided among members but never how large the pool is. And the backing of a point already held is floored rather than left to dilute, by bounding issuance against what trailing revenue can support.
The honest limits belong next to the claims. The design is built for solvency, not for price. It cannot promise what a redeemed claim will be worth in outside terms, only that the pool behind it was funded before it was spent. Steady backing per point is a design aim rather than something live today. Access has not been opened: the engine is built and running on infrastructure NOX operates, reached through Telegram as a Mini App, and the current phase is stabilization ahead of general availability. $NOX is spendable inside the catalog and is not transferable outside the ecosystem. The sections that follow mark, throughout, the boundary between what is shipped, what is specified, and what is planned.
Mission. To eliminate the financial barriers between people and advanced technology by building an ecosystem where participation creates value, and value flows back to the people who create it.
Vision. A world where opportunity isn't priced: where technology is open to everyone, and every member has the tools to build, a platform to launch on, and a stake in what they help grow.
Section: Introduction sets out the two admission prices technology ordinarily charges, the third that NOX substitutes in their place, effort in the broad sense given there, and the Proof of Clearing rule that issues reward only against revenue already cleared. This section takes those as given and turns to what follows from them.
Effort-based admission changes the shape of the transaction, not only its price. A demand-side partner funds the access, and the member is rewarded for the attention that funding buys with a claim redeemable across the catalog. Because that claim is redeemable against the same kind of products partners are trying to sell, the member who viewed an offer can become the member who acquires the product, with no upfront payment. The partner's spend does not leave the system at the impression. It funds the reward, and the reward funds a later acquisition. All three roles settle out of one cleared payment rather than at one another's expense, and the member holds a position in that settlement instead of being the good that is sold.
The design routes a member through four stages, each free to enter and none requiring that the prior stage be left behind. A member may remain at any stage indefinitely. Each rung replaces a payment ordinarily demanded at that stage with a reward or a redeemed claim: free access to tools and content, learning without tuition, trying products before owning them, and finally owning outcomes of one's own.
Most systems in this category stop at the first stage. A participant contributes attention, receives a small reward, and the loop repeats with no further step available. NOX specifies an exit from that loop. A member-originated submission that the ecosystem admits attributes 10% of its ongoing revenue to the originating member and 50% to the community reward pool, with the remaining 40% covering hosting, support and distribution. Catalog admission is decided jointly by community vote and platform curation. A participant who begins by contributing low-value effort can therefore become an attributed originator of value inside the same system, with the destination measured in ongoing revenue attribution rather than in engagement volume.
The design is built against a failure mode seen in earlier reward systems, several of which issued rewards ahead of, or independent of, matching revenue. Here, redemptions are funded exclusively from revenue that has already cleared, paid in arrears. This solvency constraint must continue to hold as the member count grows. NOX is designed against a target scale of up to one billion members, which enters this document as a design constraint on population and cost rather than as a market claim, and the model, pricing, and security work that follow are what keep the constraint structurally true at any size.
The platform gives a member direct access to a widening set of products and services, rewards the effort spent reaching them, and lets a member take paid work posted by buyers who need it done. Three modalities of rewarded participation organize that surface, and each rests on the same foundation: reward is funded only by value that has already cleared, never by newly issued tokens (Section: Proof of Clearing). A claim can be created only once the revenue behind it has settled, which is what separates a reward here from an inflationary giveaway.
The same three modalities introduced earlier organize the earning surface: access and attention, productive contribution, and routed demand. The platform's pricing of effort is most visible in the first. Rather than metering products behind a paywall, the platform opens them and rewards a member for the attention given while using them, funded by demand-side partners paying to reach that supply of effort. The tap loop is the clearest of those earning surfaces, and the rule that prices a unit of effort against it is set out in full at Section: How Rewards Work.
Productive contribution and routed demand settle into the same points ledger, and both are set out in full where the modalities are first introduced (Section: Introduction). A member who prefers not to supply attention may instead pay a subscription for ad-free access, which contributes cleared revenue in its own right.
Effort accrues off-chain in points. A member converts points into $NOX, the settlement unit inside the ecosystem, at the rate published for the current window with a level-based bonus applied, net of a 3% conversion fee. The conversion runs one way only, with no path from $NOX back to points, which is part of what keeps the ledger's solvency intact. A higher rank compounds twice, yielding more points per unit of effort and more $NOX per point converted. $NOX is spendable inside the catalog and carries no value outside the ecosystem.
Converted $NOX is a discretionary claim a member spends across a catalog of products and services. The catalog is deliberately a mutable business layer above the engine, not a fixed part of the protocol: categories and items can be added, retired, reweighted, or repriced without any change to how the engine prices effort or settles a claim. Present categories span first-party AI tools and education, in-ecosystem apps and games, and partner-supplied software subscriptions, connectivity, consumer vouchers, and financial products such as funded-trading evaluations, the last group fulfilled by licensed partners under their own terms. The catalog is also open to member-originated products, on the attribution terms and admission process set out in Section: Governance & Risk. Because the catalog floats above the ledger, a category can be restricted or withdrawn in response to a regulatory change without touching the mechanics beneath it.
One boundary governs every partner-fulfilled category. NOX distributes and introduces; partner firms fulfil. Where a catalog entry involves a regulated activity, that activity is performed end to end by the partner under the partner's own licenses, onboarding, due diligence, and disclosures, and the member contracts with that partner rather than with NOX. Regulated categories are designed to be geo-gated, with eligibility checked at the moment of redemption against the rules then in force. These controls ship with the categories they govern, and no regulated category opens before them.
Progression runs through the Neural Ascent, twenty levels in six tiers, each level bound to a distinct collectible artifact. Advancing is a burn, not a purchase: minting the next level permanently destroys the required points from the member's balance and from outstanding supply, with the full accounting of that sink set out later (Section: The Neural Ascent). Each artifact is generated deterministically from a committed seed, so the exact image can be reproduced and checked. Seeds are recorded off-chain today, and would be committed on a settlement layer only if one is ever adopted. Level is more than status: it raises the earn and conversion bonuses above, and grants a standing claim on a larger share of the reward pool, a rank that persists rather than resetting.
Member access is not yet open. The engine described here is built and running server-side, in stabilization ahead of general availability, so this platform surface reflects the current design rather than a history of live traffic.
The protocol runs on an off-chain unit called points. A member earns points for verified contribution, holds them in a server-side ledger, spends them on progression, and eventually converts them one way into $NOX. Points carry no value outside the app, and no member device can create one: only the server issues points, and it issues them only against revenue that has already settled. That governing rule is what this document names Proof of Clearing. Every reward the engine grants traces back to value a third party has already paid, never to newly minted token supply. The protocol sets a member-reward share of that cleared revenue on each earning surface, repriced against realized economics rather than fixed by assumption.
Earning falls into the three participation modalities introduced in Section: Introduction, namely access and attention, productive contribution, and routed demand. They differ in what a member supplies and in who funds the reward, and they are alike in that every one settles against value that has already cleared. The first two settle into the same points ledger; routed demand settles there as a rebate or outside it as a lower price paid. A member may also pay a subscription for ad-free access in place of supplying attention. Because a subscriber consumes no rewarded inventory, that fee clears as revenue with no points credited against it, so it adds to the reward pool rather than competing with the members who do supply attention.
The clearest illustration of how the engine prices a unit of effort is the tap loop. Each batch of taps is scored server-side by a single formula:
points = round(10 × (1 + earnBonus)) × (2 during Happy Hour, else 1)The base of 10 is the value published for the current repricing window, not a permanent constant. It is uniform, identical for every member inside a window, and repriced across windows against cleared revenue. Uniformity is the fairness property; repricing is the solvency property. earnBonus is a percentage attached to the member's progression level, rising from 0% at the first rank to 50% at the top. Happy Hour is a deterministic function of the clock, running the first stretch of every third hour, so client and server compute the same state without a synchronization call and it doubles the whole result while active. Rounding takes the nearest whole point; the protocol never rewards a fraction.
Earning across every surface is rate-limited and server-authoritative. A separate throttle, regenerating at a steady server-side rate, bounds how much any account can earn in a given span. A device reports activity, but the server verifies and meters it before crediting a point, so the tap loop cannot be minted on the client.
Points are spent to climb. Progression, called the Neural Ascent, runs twenty levels across six tiers, and each level is minted from points. Every mint burns those points permanently: there is no un-mint and no path back into the ledger, and the cumulative cost of the full climb is set out in Section: The Neural Ascent. Climbing is not cosmetic, because both the earn bonus and the conversion bonus scale with level, so a higher rank compounds twice, more points per unit of effort and more $NOX per point converted.
Points convert into $NOX, and $NOX never converts back:
$NOX = (points ÷ 1,000) × (1 + convBonus)At the rate published for the current window, 1,000 points convert to 1 $NOX before fees, and more once the account has climbed. The rate is a published parameter rather than a permanent constant, repriced against realized economics and against remaining supply. A flat 3% conversion fee is then taken from that amount as protocol revenue, so the member receives the remainder. Because the bridge runs one way, points cannot be shuttled back and forth to farm a spread, and every unit of $NOX in circulation is backed by points that actually left the ledger. Multiple surfaces put points in; two deliberate exits take them out, namely minting into progression and one-way conversion, so the ledger nets toward its sinks by construction.
Several properties follow from funding reward in arrears, chief among them that rewards are funded only by cleared revenue and never by emission, that the reward pool cannot go negative, and that a floor is defended under the backing of points a member already holds. These properties, together with the honest limits that sit beside them, are stated in full in Section: Proof of Clearing and proved formally in the NOX Technical Specification, rather than repeated here.
Most reward points are a number on a screen backed by a promise. A point in NOX is defined the other way around. It is a live claim on revenue that has already cleared, priced against money an advertiser or partner has actually paid rather than against a projection, a marketing budget, or a token that the protocol printed to cover its own obligations. The name for this arrangement is Proof of Clearing, and this section explains what it means and how it works. Two things follow, and the difference between them is worth stating rather than blurring. Where a member completes a task funded by a buyer, that buyer's payment is held against that task before the work starts, so the money exists before the point does. Everywhere else a point is a claim on a pool that has already been funded by revenue that settled, which is not the same as a deposit set aside in a member's name. Points are not money, are not held on deposit, and are not redeemable for cash. They are a claim on value that arrived, which is a different and considerably stronger thing than a promise that value will arrive.
Behind every earning surface in NOX sits one quantity that decides what a point is worth. The protocol calls it the value function, written V(t), and its logic is deliberately plain. Take the revenue an engine has cleared over a recent window. Set aside the community share, the portion of that cleared revenue allocated to members for that window. Divide that share across all the points currently outstanding for that engine. The result is the backing behind a single point.
Two consequences follow from writing it this way, and both matter. The first is that nothing outside real revenue can lift the number. A good week of downloads, a burst of attention, an announcement, none of it moves V(t), because the only input is cash that has already cleared. The second is subtler. Multiply the value of a point by the number of points outstanding and the outstanding count cancels out. The total value backing a pool is fixed by revenue alone. Issuing more points does not create more backing. It divides the same total into thinner slices. Retiring points, which happens every time a member burns them to climb a level, removes slices without touching the pool, so the value of every point that remains rises.
The technical specification proves a set of properties about this system formally, by working through the exact rules the server enforces on every ledger write. The proofs are not reproduced here. What follows are the guarantees themselves, in plain language, so that a reader can hold the protocol to them without reading the algebra. Each one is proved formally in the NOX Technical Specification.
There is one recurring way money enters the reward pool: an advertiser or partner remits payment, it settles, and the community share of that settled amount is credited. A booked invoice credits nothing. A pending commission credits nothing. The pool also carries a founding seed contributed by the operating company at launch, and first-party products enter the catalog at their cost to NOX. Critically, no reward is ever funded by creating tokens. There is no mechanism anywhere in the system that mints value to cover a payout. Every reward a member can draw traces back to revenue that already arrived. This is the property that separates Proof of Clearing from the arrangements where early participants are rewarded out of the deposits of later ones.
The pool is credited only when revenue clears and debited only when a redemption actually settles, and a redemption is permitted only when the pool holds enough to cover it. Money out is bounded by money in, because that is how the pool is defined rather than because the operator undertakes to keep it so. Rewards drawn in a period are bounded by the community share of revenue cleared for that period. Solvency here is a property of the accounting rather than a matter of trust, and like any accounting property it holds only so long as the rules above are the rules actually run.
Not every member draws the same share. A personal contribution weight, built from a member's permanent Ascent level, recent activity, and what they have brought to the network as a builder, scales how large a slice each member receives. But weight only decides who draws more of a fixed total. It cannot make the total larger. A member with a large weight advantage is receiving a bigger portion of the same pie, not a pie that grew to accommodate them.
This property protects the pool. It does not, on its own, protect an individual member. Because the total is fixed, a weight inflated by manipulation is not funded by the operator. It is funded out of the shares of honest members. That is why the enforcement that keeps weights honest belongs to the mechanism itself, and it is described where the server's authority over the ledger is set out. The protocol is server-authoritative, earning is rate-limited so that no account can manufacture activity faster than the server grants it, behavior is scored, and accounts that cross into abuse are limited. The design is built to keep manipulation bounded, because zero-sum redistribution makes honest members, not the company, the ones a manipulator would be stealing from.
Solvency and value are two different properties, and it is worth being honest that a system can hold the first while quietly failing the second. The pool can stay solvent, refusing no redemption, even while the value of a point a member already holds declines, if new points are issued faster than revenue grows. Holding only solvency protects the operator and leaves the member exposed.
The design therefore aims at a second property alongside solvency: keeping the backing behind a point steady as new points are issued, rather than letting issuance outrun the revenue behind it. The floor is not defended by refusing to reward effort. Every verified action settles immediately at the rate published for the current window, and no credit already earned is ever revoked. Instead the floor is defended by adjusting the size of each new grant. When revenue contracts, the next grant per action shrinks. As members burn points at level-ups, room reopens and grants recover. The floor moves with realized economics rather than with a figure fixed in advance.
One honest caveat belongs here. This is design intent rather than something a member can observe today. It describes what the system is built to aim at, and it is not yet live. In the same spirit, member access is still being stabilized ahead of general availability.
Everything else NOX offers, the tasks, the progression, the token, rests on a single claim: that the rewards are real, in the specific sense that they are backed by revenue that has already cleared and never by value the protocol invented to pay itself. Proof of Clearing is that claim made precise. A point is a slice of money that arrived. The pool cannot promise more than arrived. Weight decides who gets which slice, never how much there is to slice. And the backing under a point a member holds is what the issuance rule is built to keep steady. These are the properties the rest of the design is accountable to.
The Neural Ascent is NOX's progression layer, and it is also the economy's largest deliberate sink. It is a ladder of twenty levels organized into six tiers, running from Dusk at the base to Aurum at the summit. Each level is bound to a distinct collectible artifact. The Ascent accepts a single input: points, issued by the engine's contribution-weighted mechanism and then burned by the member who chooses to climb.
A level is more than a score. It is the artifact a member holds and the rank shown against their name, and that identity is auditable. Every artifact is generated from a seed, and it is the seed that is recorded rather than the finished image. Anyone can rerun the published renderer against the seed and reproduce the exact scene. Rarity is therefore something anyone can check by reproduction, not a claim the issuer asserts.
Advancing a level is an irreversible burn, not a purchase. Once a member's point balance meets the cost of the next level, minting destroys that quantity of points. The amount is subtracted permanently from the member's balance and from the outstanding supply the economy tracks. For a member who climbs the full ladder, minting becomes the single largest permanent point sink of their time on the platform,.
Σ burn (levels 1 to 20) = 2,172,000 points, from 0 at the first level to 420,000 at the twentiethHolding all twenty artifacts costs 2,172,000 points, permanently retired from supply. The cost grows disproportionately with each level, so scarcity at the summit reflects points actually burned to reach it rather than a rarity value assigned from above.
Minting issues no new points, refunds nothing, and offers no volume discount. It only destroys. Each level a member mints reduces outstanding supply, which raises the value of every point still held by every other member, all else equal. Progression through the ladder is a working part of the point-supply deflation the pricing model depends on, not a status display bolted onto the side of it.
The level a member holds raises their tap and conversion bonuses, and it does one thing more: it feeds the contribution weight that decides the member's slice of the pool, which is the same thing as what their points are worth. The contribution-weighted pricing model recomputes each member's weight weekly from three terms, and one of those terms is the level a member has reached.
The level term ranges from one at Dusk, on day one, up to three at Aurum, the full climb. Unlike the activity term, which decays after an idle week, the level term records permanent earned rank and does not reset. It is a standing claim on a larger share of the pool, earned once and kept. The fairness properties of an earned-only ladder are proved formally in the NOX Technical Specification.
Inside the app, members earn points, and a point is a private ledger entry. The company that keeps the ledger can in principle rewrite it, dilute it, or freeze it, and the holder has no independent way to check. NOX already runs its points on a ledger that is server-authoritative, audited, and resistant to manipulation, but that ledger is private, and a private ledger, however carefully kept, remains a promise from one company that only that company can verify. $NOX is the mechanism that turns the promise into something a member can check for themselves.
Four properties motivate a token rather than a larger points system. Its total supply is auditable, so scarcity can be checked rather than taken on faith. Ownership becomes independent of the issuer once the token sits in a member's own wallet. Scarcity is tied to real usage, because redemption and level advancement remove supply while nothing is issued for mere presence. And distribution begins only after the mechanisms that create demand have already been running inside a points-only app.
NOX runs two ledgers, kept deliberately apart. Points are the off-chain earning unit, rewarded server-side at published rates and carrying no value outside the app. $NOX is the settlement unit inside the ecosystem. It is credited on conversion and is spendable across the catalog from launch. At launch it is closed-loop store credit: it is not transferable between members, is not listed on any market, and is not redeemable for cash. Moving $NOX onto a settlement layer of its own is an aim rather than a schedule, described in Section: Chain Architecture. The bridge between the two runs in one direction only: points convert into $NOX, and $NOX never converts back into points.
That one-way design is structural rather than cosmetic. It prevents a token bought on an exchange from being presented inside the app as if it had been earned, and it keeps the token's supply grounded in contribution, which cannot be purchased directly. The conversion itself is straightforward.
NOX_out = (P / 1,000) × (1 + β_L)
P is the number of points a member converts. β_L is a bonus attached to the member's Neural Ascent level, starting at zero at level 1 and rising through the tiers to 30% at level 20. The level bonus therefore enters the exchange rate at the single point where value leaves the points economy. A separate conversion fee of 3% is taken from the credited $NOX as company revenue, kept outside the exchange rate rather than folded into it.
The working model sets a fixed total supply of 1,000,000,000 $NOX, allocated as below. Both the total and the allocation are published here as illustrative, and are confirmed and disclosed before TGE. The contract is mint-closed after genesis except for the one-way conversion bridge, so the total cannot be inflated and outstanding supply can be checked on-chain against this figure.
| Allocation | Share | Purpose |
|---|---|---|
| Rewards pool | 87% | Funds the points-to-$NOX bridge as members convert, released on usage rather than on a timer |
| Liquidity | 5% | Reserved against any future settlement layer |
| Team | 5% | Vested, on standard founder and employee alignment terms |
| SAFT / pre-sale | 3% | Reserved for early financing on terms set out in definitive documents |
The large share assigned to the rewards pool follows from a single design goal: supply should enter circulation through member activity rather than through distribution to insiders, released as conversions occur rather than on a fixed timetable.
Two distinct pools of $NOX exist, and the relationship between them is worth stating plainly rather than leaving to inference. Any listed token needs supply that member activity cannot produce: exchange liquidity, market-making inventory, and team allocations cannot be brought into existence by conversions. That supply is real, is disclosed in the table above, and is minted once at genesis. The property that matters is narrower, and stronger, than its absence would be. No allocation outside the rewards pool has any path into the member reward pool, and none of it can therefore fund a member reward.
The separation holds on both ledgers rather than resting on a promise. On the points ledger, the reward pool accepts exactly one kind of credit, funded by cleared revenue and by nothing else, so no token balance can be routed into it. On the token ledger, nothing is minted after genesis except through the one-way bridge, so the rewards supply grows only as points are destroyed to create it. Member rewards are therefore never funded by token supply, whether that supply was granted at genesis or minted through the bridge. This property is proved formally in the NOX Technical Specification.
Earlier token designs of this kind tend to share one failure: a faucet that never closes, issuing new supply on a schedule until the price collapses under it. NOX uses sinks instead, mechanisms that permanently remove supply as members use the product, and it runs two of them on two different ledgers.
Both sinks are driven by the same underlying variable, which is product usage, and no schedule rewards holding or staking before a product exists for the token to be spent against. The 3% taken on each conversion is protocol revenue rather than a burn, and as revenue it enters the pool calculation on the same terms as every other cleared receipt.
Revenue recycling. A share of cleared revenue is directed back to members through the community reward pool. It is funded from income already collected and never from issuing additional token. Were $NOX ever to become transferable, buying supply back on the open market would be one further way to return value, but no such market exists at launch and none is undertaken here. The community share is taken from the same cleared receipt as operating costs rather than from what is left after them.
At launch $NOX is closed-loop store credit issued by the operating company. It is spendable inside the NOX catalog, and it is not transferable, not listed, and not redeemable for cash. No token has been distributed on any public network. The classification of $NOX has not been settled, any classification is jurisdiction-specific, and it would be confirmed with counsel and, where appropriate, with the competent regulator before anything is distributed. Distribution would follow product deployment rather than precede it.
This document does not constitute an offer to sell securities or digital assets, or a solicitation of an offer to buy, in any jurisdiction. Any future distribution of $NOX will be governed by definitive legal documentation and is subject to applicable law, including securities and digital-asset regulation where relevant.
The economy has one hard rule, and every other business decision follows from it. A reward is issued only against revenue that has already cleared, never out of an expectation, a projection, or freshly minted token, so a member's balance grows only after real money has arrived from a partner or a subscriber and only a share of that arrived money goes to rewards. The properties that follow from funding reward in arrears, among them that the reward pool cannot go negative, are set out in full at Section: Proof of Clearing.
The reward budget is a defined share of the revenue that clears, set for each earning surface and repriced against realized economics rather than fixed in advance. The remainder covers infrastructure and operations. Because the budget is a share of what has already been collected rather than a fixed sum promised ahead of it, it moves with cleared revenue. When revenue is thin, rewards are smaller and the pool stays solvent; the design never borrows from future clearing to flatter present payouts.
Revenue reaches the ecosystem through four structurally different channels. The distinction matters because each carries a different margin, a different degree of NOX control, and a different regulatory posture, and because no single channel gates the others. They clear on different schedules and scale on different variables, so a shortfall in one does not cascade into a payout failure in another.
The first channel is first-party revenue, from products NOX builds, owns, and operates directly. This includes its own AI tools, education offerings, games and engagement surfaces, in-ecosystem mini-apps, an optional ad-free subscription whose cleared fee adds to the pool rather than competing with it (Section: How Rewards Work), and a first-party rewarded ad network. No intermediary sits between NOX and the member here, so control over pricing, availability, and integration is complete, and this is the tier NOX invests in over time.
The second channel is distribution and partnership revenue, earned by introducing third-party products through the catalog. Software and AI subscriptions, connectivity, consumer vouchers, and financial products are supplied end to end by partner firms under those firms' own licenses, terms, onboarding, customer due diligence, and disclosures. NOX earns a commission on activations and, where a partner's terms allow, a revenue share on the customers it originates. Financial-product access is geo-gated, and the perimeter is firm across the whole tier: NOX distributes and introduces, and does not itself conduct any regulated financial activity.
The third channel is protocol revenue, generated by the mechanics of the ecosystem as members use it. This includes a modest margin on catalog fulfilment, commissions on in-ecosystem purchases, the 3% fee applied when points are converted to $NOX, and, from token launch, minting and marketplace fees. This tier needs no external partner and carries no licensing surface. It grows with the two things the whole design is engineered to grow: verified activity and successful redemption.
The sequencing is deliberate. Partner distribution monetizes earliest, because partner catalogs exist before first-party ones do. First-party products are the strategic destination, and over time demand shifts toward what NOX builds itself, compounding margin and control without ever bringing a regulated activity in-house. NOX intends to report the changing mix across these four channels, because the mix, more than any single stream, shows whether the ecosystem thesis is being executed.
Specified, not yet live. The member side of task completion is built. The buyer side, point purchase, escrow and dispute handling are designed and not yet in production.
The fourth channel is the one that makes NOX a marketplace rather than a rewards app. Advertising sells a member's attention to a buyer who wants it seen. This channel sells something narrower and more valuable: a specific action, completed by a verified person, proved before it is paid for. A buyer funds a task pool with ordinary money, defines what completing the task means, and members who complete it are credited from that pool.
What a buyer can ask for spans a wide range, and the range is the point. A web3 project can ask members to join a community channel, follow an account, or test a release before launch. A brand can commission content, reviews or translations in a language and a region where it has no presence. A business can ask for something physical, whether that is verifying that a shop display was installed, photographing a location, standing outside a venue with a sign, or running an errand. A company that needs freelance work done can post it and hire from a pool of members whose reliability is already recorded. As machine buyers become able to commission work, the same mechanism lets software hire people for the parts of a job that still need hands, eyes or a face.
Three properties make this channel structurally different from the others. The money arrives before the work is done rather than after, which is the cleanest possible fit with the rule that rewards are funded only by revenue that has cleared. The buyer is not paying for the chance that a member notices something, so the value of a completion is set by agreement rather than by an auction NOX does not control. And a member's standing is directly useful here in a way it is not on an advertising surface, because a buyer choosing who completes their task cares a great deal about who has completed tasks reliably before.
Funds are held until completion is evidenced, and what counts as evidence depends on what was asked. Joining a channel is checkable against the platform's own interface. Content and translation are reviewed before release. A physical task is evidenced by what the buyer specified when the task was posted. Where the two sides disagree, the funds stay held until the dispute is resolved, and neither side can move them unilaterally.
NOX earns on this channel in two places. A fee is taken when a buyer purchases points, and the same 3% is taken when a member converts points into $NOX. The economics are deliberately thin on each transaction and rest on volume, because the purpose of the channel is to bring buyers into the ecosystem who would never have bought an impression.
The three ways a member participates, access and attention, productive contribution, and routed demand, are set out in full at Section: Introduction. What matters to the business model is that all three draw from the same reward pool: whichever way a member earns, the credit comes from cleared revenue under the same pool rule, never from a separate budget and never from emission.
The core idea closes a loop. When a partner already lists a product in the catalog and buys rewarded attention inside NOX, the money spent clears first, and then a share of it is issued to the members who watched, denominated in the same catalog where that product sits. Advertising spend that a conventional network keeps as margin is redirected into pre-funded demand held by the people most likely to redeem it. Around this sits the recycling of cleared revenue back into the ecosystem through the reward pool and, after TGE, buybacks, as Section: The $NOX Token sets out in full.
The reward mechanism described earlier holds without reference to any blockchain, because every rule it relies on is enforced against a server-side ledger. The settlement chain sits at the same level of the design: a component of the protocol, not a foundation it is built on. The points ledger, the solvency rule and the redemption logic are each defined independently of any chain, so chain selection can be treated as a choice made against published criteria rather than a technology adopted on faith. Deployment is described here in phases, and one constraint holds across all of them: no member is ever required to fund a wallet before earning or redeeming a reward. The later phases describe an ambition for the ecosystem rather than a commitment to a date or an outcome.
In the current phase NOX operates with no on-chain dependency. Every contribution credits an off-chain, contribution-weighted points ledger held server-side, and the server is the only party able to create or move a point. No blockchain call appears in the earning or redemption logic, and chain selection is deliberately deferred until it can be made against verified information rather than under the pressure of onboarding speed. Points remain non-transferable and are shown against $NOX for display only, so no member balance is committed to a settlement layer while the landscape is still being observed.
Were $NOX ever to move onto a public settlement layer, it would become a transparent, auditable on-chain asset, and chain selection would follow a criteria-first procedure: the evaluation criteria and their weights are defined before any candidate chain is scored, so the record of the method precedes the record of the decision. The criteria weigh gasless experience, Telegram-native distribution, fee economics, wallet reach, compliance and decentralization trajectory. On the information available today TON leads, with an EVM Layer 2 as an evaluated fallback. No selection has been made and no date is set.
The engine never calls a blockchain directly. Every mint, transfer, burn and balance check passes through an internal adapter, behind which sits the chain-specific implementation, and any one implementation can be swapped for another without touching the earning or redemption logic. This carries the Phase 1 no-lock-in property forward and keeps the chain an interchangeable detail rather than a structural dependency.
The ambition, once the ecosystem is stable enough to carry it, is a settlement layer operated by NOX, on which $NOX functions as the utility unit and transactions stay gasless for members. Whether that is ever built, and in what form, is undecided. One rule would govern the design, stated plainly because this is where a settlement layer could quietly undo the rest of the document.
Whether NOX ever operates its own settlement layer is undecided and outside the scope of this document. No staking, yield, or revenue-sharing arrangement of any kind is offered or contemplated here.
A security budget bought with token issuance would reintroduce at the settlement layer the exact failure the reward mechanism is built to exclude, and no gain in sovereignty would be worth that trade. Nothing in this document offers or contemplates staking, yield, or any revenue-sharing arrangement. Because the same adapter fronts any implementation, a later change of settlement layer would be a substitution rather than a rebuild.
No member spends gas to use the engine. In Phase 1 there is no gas, because the ledger is off-chain. In Phase 2, the gasless capability on TON is governed by TON itself rather than by NOX, so NOX sponsors transactions as a second layer and covers directly any residual cost the network does not absorb; under an EVM fallback, any residual gas cost is met by NOX. Were a NOX settlement layer ever to exist, the same intent would carry into it. A member who has never held a cryptocurrency balance can mint an artifact, claim a level or redeem a reward without funding a wallet, at every stage. The settlement layer is present in the architecture but never exposed as a precondition for taking part.
The reputation layer is not introduced alongside a settlement layer. It operates from the moment a member first earns. The same behavioral signal that scores fair play extends, as the engine moves on-chain, to gas-sponsorship eligibility, earn-rate multipliers, payout caps and catalog access, so a single ranking mechanism determines both community standing and on-chain standing.
A rewards economy is only as trustworthy as the machine that computes it. NOX starts from one refusal that shapes everything else: the client never computes a balance, and the server never trusts a number the client sends. The interface a member sees renders state and reports intent. The server recomputes every reward from data it owns and credits the member accordingly. A modified app that claims a thousand actions against the effort for forty is credited for forty. There is no path in which the number on the device becomes the truth.
From that footing a few plain properties follow. Earning is rate-limited, so no account can manufacture activity faster than the server is willing to grant it. Behavior is scored from signals the server can see, and accounts that cross into abuse have their earning bounded rather than left to run. The mechanics that do this, the exact limits, and the classes of attacker they are measured against are set out in the NOX Technical Specification and are deliberately not reproduced here. What matters for a reader is the shape of the mechanism, not its wiring.
The design is provably abuse-bounded, and the proof is carried out formally in the technical specification. Stated plainly, the property is this. Because the reward pool is set by cleared revenue and shared across outstanding points, no coordinated group of accounts can extract more than the pool already holds. A failure of detection can change who receives value, moving it away from honest members, which is exactly why anti-abuse is a first-class part of the system. It cannot manufacture value the pool was never funded for. Manipulation is contained inside a total that adversarial pressure cannot enlarge.
The handling of a flagged account is built to be reversible. Earnings under review are held aside, not erased, and each case is examined on its own facts rather than run against a clock. Because held earnings are never forfeited while a review is open, taking the time to reach the right answer costs the member nothing.
Illustrative. The rates below are the ones the model is built on and they move with realized economics, so they are shown to make the shape concrete rather than as a fixed promise.
Take the plainest case: a member who only earns from the tap loop and does nothing else. Five taps bank a round and an ad seals it, so a round is one completed ad. Around two thousand rounds earn enough to take a subscription to one of NOX's own products listing at $20, which is the kind of reward the catalog is built around because NOX supplies it at its own cost rather than buying it at retail.
Two thousand rounds is not a small number, and it is worth being straight about what it means. At a typical eight rounds a day it is a little over eight months, and at the daily limit the product enforces it is under a month. Those limits exist because an unlimited counter is a bot magnet, not because the pool cannot afford more.
The same arithmetic runs downward, which is the part that matters more day to day. If two thousand rounds reach a $20 subscription, then a hundred reach something worth a dollar, which is a matter of days rather than months. A member does not have to save toward one large reward. Most of the catalog sits well below the price of a subscription precisely so that the first redemption arrives early rather than at the end of a quarter.
Two things are deliberately absent from this example. It assumes no progression, and a member who has climbed the Ascent earns more for the same round, up to half as much again at the top. And it assumes only one surface, when the same member is also earning from tasks, referrals, routed play and, once it opens, work posted by buyers in the marketplace. The tap loop is the slowest way to earn in NOX, which makes it the honest one to quote.
These are the terms as they stand today. Where a question is not yet settled, it is marked as such rather than answered provisionally, and the full terms of service govern in every case.
Balances. Points do not expire. There is no inactivity forfeiture and no timer running against a balance a member has earned. The smallest amount that can be converted from points into $NOX is 100 points.
What $NOX is, and is not. $NOX is closed-loop store credit. It is spendable inside the NOX catalog, it is not transferable to another member, it is not listed on any market, and it cannot be exchanged for cash. A member who stops using NOX does not hold an asset they can sell.
Enforcement and appeal. Where an account is limited or held, the decision is recorded against the specific named signal that caused it, and a member may ask for it to be reviewed by a person. An account found to have been limited wrongly is restored, and the earnings held during the review are released with it.
Eligibility. NOX is open worldwide, other than in jurisdictions subject to sanctions or where the law does not permit it. NOX is not intended for children, and some categories, particularly anything financial or regulated, are unavailable in some countries. Which categories are available where is determined by the rules in force in each place and can change.
Data. NOX records what an account does inside the product, including device and session information, because that is what makes it possible to tell a member apart from an automated account. Raw health data from wearable engines never leaves the device.
Not yet settled. The following are undecided and will be published before they apply: the minimum age by jurisdiction, the period for which behavioural records are retained, the time within which a review of an enforcement decision is answered, and what happens to outstanding balances if NOX ceases to operate. None of these are deferred because the answer is unfavourable. They are deferred because they depend on advice not yet taken, and stating them early would be worth less than stating them correctly.
Specified, not yet live. The enforcement side described at the end of this subsection is built and running today. Standing itself is designed and not yet in production.
An account that has been present for months, returns without prompting, completes what it starts and spends what it earns is worth more to the ecosystem than one that arrived this morning. NOX records that difference as standing. Standing is not bought and it is not granted. It accumulates from ordinary use over time, and it is the ecosystem's way of telling a long-standing member apart from an account that was created an hour ago to extract value and leave.
Standing draws on the things a member does rather than on anything a member declares. How long the account has been active. Whether it returns across weeks rather than in a single burst. Whether tasks that are started are finished. Whether earned balances are actually spent in the catalog. Whether the members it refers go on to become real participants themselves. Each of these is expensive for an automated account to imitate, because each of them costs the one thing a bot farm cannot manufacture cheaply, which is sustained, genuine, human time.
The aim is that standing shapes what an account earns from the same activity, so that the members who build the community are the members the community rewards. The intention is not to punish a member who misses a week. Ordinary life interrupts ordinary use, and standing is designed to reflect a long pattern rather than a perfect one, so that participating in NOX does not become an obligation.
Standing has a counterpart. The same server-side record that recognises genuine participation also recognises its opposite. Where the evidence shows an account is automated, or is one of many operated by a single party to multiply a single person's share, its earnings are held rather than paid, and the value that would have gone to it stays in the community reward pool where the members who earned honestly draw from it. Value is not taken from one member and handed to another by name. It simply never leaves the pool, which is where every member's rewards come from in the first place.
Two commitments govern how that power is used, because a mechanism that can withhold earnings is a mechanism that has to be accountable. Every enforcement decision is recorded against the specific named signal that triggered it, so that a decision can be reviewed by a person rather than defended as the output of a model. And an account that is found to have been treated wrongly is restored. The purpose of the mechanism is to protect the members who participate honestly, not to reduce what is paid out.
NOX runs on data, and it is honest about that rather than quiet. Historically, consumer software has been funded by data whose value was never disclosed to, or priced for, the person generating it. NOX makes the exchange explicit. Before an action is taken, the interface states that it will be recorded and what it earns. Collection is disclosed at the moment of the action, not buried in a settings page, and the data that funds the reward pool is auditable by construction rather than inferred from network traffic.
The opt-in structure is architectural, not an addendum. Each category of data sits behind its own granular consent, and an optional category can be revoked without any loss of earning access. Personally identifying data is not sold or transferred as a resellable asset; what moves downstream is aggregate or pseudonymous. Identity verification is not required to earn or to hold points. It is deferred to a compliance gate ahead of real-value redemption, on a simple principle: identity that was never collected cannot be exposed. Research participation runs its own separate, independently revocable consent flow.
Specified, not yet live. These engines are designed and not yet in production, and the boundary described here is what they are being built to. The Sleep and Exercise engines are opt-in and are designed to hold the strictest boundary in the product. Raw health data never leaves the device. Heart rate, step-by-step motion, and sleep-stage detail are not uploaded, stored, or transmitted. What reaches the server is a completion proof: a compact signal that a threshold was met, and nothing more. This is the one surface where the server cannot recompute the underlying evidence for itself, so the proof is bound to the device and the platform health API that produced it, and these engines carry tighter caps than surfaces the server can verify directly. The reward is issued without the biometric stream that established the fact ever being held. This is not a policy layered over a pipeline that could carry more. The pipeline was built without the capacity to receive raw wearable data in the first place.
One completion event then serves three parties, none holding more than its role requires: the member, rewarded for a documented action; the sponsor, given a verified outcome; and the separately consented research panel, given identity-stripped signal.
The through-line across security and privacy is the same. The server is authoritative, so a reward is trustworthy. Consent is granular and revocable, so the data behind it is too. And the least data that can prove a fact is the most the system is built to hold.
Governance extends a member's authority from the data they generate to the protocol itself, and it arrives before any token exists to gate it. An in-app proposal-and-voting system has already been built and deployed, reached through Telegram as a Mini App, and is ready for use once member access opens. Members submit proposals, the community deliberates, and the proposals that accumulate the strongest support inform the near-term product backlog. Voting weight is earned through participation rather than purchased, so governance influence is decoupled from capital from the outset.
Catalog curation runs a parallel pipeline: BUILD → SUBMIT → INTEGRATE → FOUND. The community votes on which submissions merit a permanent place in the catalog, and the platform curates on top of that vote for security, quality, and fit. A submission that clears both the community vote and the platform review is integrated and exposed to the full member base. Authority is split rather than shared: the community admits, the platform constrains, and neither integrates a submission alone.
The builder arrangement is fixed and uniform across submissions. The originating builder retains 10% of the ongoing revenue of what is integrated into the catalog, 50% is credited to the community reward pool, and the platform retains 40% to fund the integration, distribution and ongoing operation of the resulting product. The split does not vary by deal size or by negotiating position. It is identical for the first builder and the thousandth, which is what lets a member move from playing on the platform to owning a product distributed across it.
NOX is developed and operated by EduChain Limited, which is the entity behind this document and the counterparty to every commitment described in it. A separate holding company to acquire and hold the technology and the intellectual property in it is planned but not yet incorporated, and nothing in this document should be read as an offer relating to that company.
Nothing in this whitepaper is an offer or solicitation to buy securities or digital assets in any jurisdiction where such an offer would be unlawful. Any subscription for shares, or any agreement relating to $NOX, is governed exclusively by definitive legal documents negotiated directly with investors by EduChain Limited. This document is descriptive and not contractual. Where the two conflict, the definitive documents control.
$NOX does not exist as a tradable token as of this writing. It is displayed in-app for planning purposes only, ahead of a future Token Generation Event. Supply allocation, points-to-$NOX conversion mechanics, and holder-tier benefits described in this document are pre-TGE proposals, subject to change ahead of launch. Digital assets are volatile, may be illiquid, and may be restricted or reclassified by regulators in any given jurisdiction, including after launch. Nothing here should be read as a promise of future value, liquidity, or exchange listing.
Catalog products are supplied by third-party partner firms under those firms' own licenses, terms, and risk disclosures. NOX operates a distribution and rewards layer here. It is not the counterparty to a funded account or a leveraged position, and it provides no investment advice. NOX further depends on infrastructure it does not control, and unilateral changes by any such partner can affect its product, economics, or timelines. The NOX Technical Specification carries the full disclosure, including leveraged-product suitability and the medical-advice disclaimer covering wellness surfaces.
The NOX Technical Specification documents failure cases in which emissions-funded reward loops collapsed once reward supply outpaced any offsetting sink. NOX's design addresses that specific failure mode. Contribution-weighted point pricing floats point value on cleared revenue rather than on emissions, and a solvency rule bounds redemptions by cleared revenue at every point in time, a property proved formally in the NOX Technical Specification. That construction removes the unbacked-emissions failure mode common to those cases, but it remains a design property under ongoing operation, not a guarantee of outcome. Execution risk, ad-market conditions, and competitive dynamics can still affect results.
NOX's reward economics are built to reprice automatically with realized ad rates rather than to promise fixed payouts. The advertising rates and affiliate commissions the reward pool depends on are set by advertiser demand, seasonality, and macro ad-spend cycles outside NOX's control. A sustained downturn in ad demand lowers earn rates for members and contribution margin for NOX at the same time, by design. The mechanism protects solvency, not growth.
Where members complete paid tasks for third-party buyers, the relationship between the buyer, the member and NOX may be characterised differently in different jurisdictions, including as engagement of a worker rather than as a reward for participation. That characterisation carries tax, employment and consumer-protection consequences that vary by country and that may change. NOX intends to operate the marketplace with terms, eligibility rules and geographic limits set with counsel, and some categories of task may be unavailable in some jurisdictions or withdrawn. Buyers are responsible for the lawfulness of what they commission, and tasks that are unlawful, unsafe, deceptive, or that ask a member to misrepresent themselves are outside what the marketplace permits.
NOX currently operates without KYC, with identity verification gated to the point at which real value leaves the system: funded-trading evaluation payouts and future redemptions of material value. Digital-asset regulation, game-mechanic rules, and rewards-app oversight are unsettled and jurisdiction-specific. The requirements NOX must meet may change before or after TGE, and that includes $NOX's eventual classification. NOX may need to restrict features or geographies in response.
Tokenomics parameters, unit-economics ranges, and other forward-looking statements in this document are forward-looking and reflect current plans and assumptions as of July 2026. They are not commitments. Actual results may differ materially, and these parameters are subject to revision as the protocol and its operating environment evolve.