The NMESH token.

What NMESH is for

NeuraMesh settles compute. A provider commits GPU capacity, a buyer or an agent reserves it, delivery is metered and verified, and payment is released on evidence. That loop produces three things a token can honestly do: pay for the risk that a provider fails, lock supply behind the right to sell capacity, and give buyers a reason to settle through the protocol rather than around it.

NMESH is designed around those three jobs and nothing else. It is not the unit providers are paid in (they receive USDC), it is not required to use the marketplace, and it carries no claim on NeuraMesh the company. Its value, if any, comes from fees that real settlement generates and from the supply those fees remove or lock.

Supply and allocation

Maximum supply is 1,000,000,000 NMESH: a Move coin (sui::coin), 9 decimals; the full supply is minted once at genesis into labelled escrow and vesting objects and the mint capability is destroyed in the same transaction. Nothing can be minted after that. Every bucket below exists from day one inside an escrow or vesting object, and the two usage-linked buckets (delivery rewards and ecosystem programs) are released only as the usage happens. Delivery rewards that are never earned are burned when their schedule ends.

Share of the 1 billion maximum supply by bucket. Hover or focus a bar for the token count.

Two design choices separate this from a typical launch table. First, the largest bucket belongs to providers and is released from escrow only against verified GPU-hours, so the supply that reaches the market is paid for with delivered compute. Second, insiders hold 30 % between contributors and backers, both behind a one-year cliff with on-chain streaming and no acceleration, which is at the conservative end of comparable networks.

Unlock and release schedule

The curve shows the most NMESH that can be in circulation in each month after launch, counting every vest at full speed and every escrowed bucket released at its ceiling. Real circulation can only be lower, and total supply falls as the burn engine and end-of-schedule escrow burns remove tokens.

Maximum circulating supply as a share of the 1 billion cap. Points mark launch and year boundaries; hover for exact values.

Net-deflation rule. For the first 24 months the delivery-reward escrow releases its schedule in full against verified GPU-hours. From month 25, only a 25 % floor of each epoch's schedule is released unconditionally; the remaining 75 % is released in proportion to the previous quarter's burn. Releases can never exceed the schedule, and whatever the escrow has not released when the schedule ends is burned, so circulating supply growth falls toward zero as settled volume grows while total supply only ever falls.

The launch float of 13 % is deliberately small but fully explained: it is the public auction, half of the community distribution, the treasury's liquidity seed and the first ecosystem budget. There are no undisclosed over-the-counter allocations and no market-maker loans that return tokens to the treasury off-schedule.

Value accrual: burn, stake, bond

The strongest token models in the market each do one thing well. Ethereum ties scarcity to usage by burning fees. BNB burns on a published formula until a hard floor. Uniswap routes protocol revenue to the people who stake. Solana tapers emissions on a predictable curve. NMESH combines all four, and adds a fifth mechanism that only a compute network can offer: providers must bond tokens to sell capacity.

From the first settled contract, net protocol fees are split 3 ways:

Why this is better than any one of the models it borrows from:

• Burn is paid for by compute, not speculation. The buy-and-burn spends fee revenue, so it only accelerates when settled GPU-hours grow. There is no treasury-funded burn theatre.
• Stakers earn revenue, not inflation. The staker share is paid in USDC. A high headline yield cannot be manufactured by printing NMESH.
• Releases are gated by burn. From month 25, 75 % of each epoch's provider rewards leave escrow only in proportion to what the burn removed, so circulating supply cannot outrun usage.
• Bonds lock supply where demand is. Every provider listing capacity bonds NMESH sized to the value committed. Supply grows with the network's capacity, and so does the amount of NMESH that cannot be sold.
• The floor is programmatic. Burning stops at 500,000,000 tokens, exactly half the cap, after which the 40 % burn share flows to stakers. Nobody has to decide when to stop.

Utility for every participant

Each use has its own rules and its own boundary, so a holder voting on policy does not automatically become an insurer, and a buyer's discount does not depend on staking.

The 6 % compute discount is the one utility that already appears across the site and the economics documentation. Its funding comes from the broker commission and the ecosystem bucket's discount reserve, never from new issuance, and it is capped per account and per budget period so it cannot be farmed by self-dealing purchases.

Staking and the provider backstop

Staking in NeuraMesh is underwriting. When a provider fails to deliver a verified contract, the failure waterfall runs in a fixed order, and staked NMESH is the layer before the treasury:

1. Buyer refund from escrow
2. Provider USDC collateral
3. Provider NMESH listing bond
4. Staked NMESH backstop (pro rata, capped)
5. Treasury reserve, by governance vote

Exposure is bounded: at most 10 % of a staker's position per incident, applied pro rata across the pool. Unstaking has a 14-day unstake cooldown; cooldown positions still cover incidents opened before the request. In exchange, stakers receive 40 % of all net protocol fees in USDC, weighted by lock tier (0, 3, 12, 24 months at 1.0, 1.2, 1.6, 2.0). Longer locks earn more because they underwrite longer-dated capacity contracts.

This is the same real-yield logic as a fee switch, with one honest addition: the yield is compensation for a risk the staker actually carries, which is what makes it sustainable without inflation.

Governance and immutable guarantees

Governance is parameter governance. Staked NMESH votes, with the same lock-tier weights as yield, and every change passes a 30-day timelock during which the parameter and its effective date are shown on this page. Some things are fixed at genesis and cannot be voted on at all:

Everything else moves only inside published bands:

How it compares

Measured against the models the market already rewards, NMESH is not a new idea in any single column. It is the first design that puts the four proven mechanisms behind one physical unit, the verified GPU-hour, and adds a supply lock that scales with the network's own capacity.

Illustrative economics

These scenarios apply the pilot broker commission from the pricing page to three levels of monthly settled volume at a flat $2 per GPU-hour. They show how fees flow, not what NMESH is worth; they are not forecasts, and they ignore trading and settlement fees entirely.

The point of the table is the shape, not the size: every dollar of fee does three things at once, and two of them (burning and paying stakers) make holding NMESH more attractive exactly when the network is used more. Volume that does not settle through the protocol does none of this, which is why the discount and the listing bond exist.

Launch path to January 2027

1. Q4 2026
   Testnet points and auditsNon-transferable testnet points record verified provider and buyer activity for the community distribution. Coin, escrow, vesting, bond and backstop Move modules go to independent audit. Legal classification opinion obtained for launch jurisdictions.
2. December 2026
   Parameters frozen and publishedFinal allocation, vesting streams, fee split and bond sizing are published here and in the docs. Escrow and allocation objects are created and labelled on-chain. Contract addresses appear on the official-links page before anything is tradable.
3. January 2027 (target)
   Token generation eventGenesis mint into escrow and vesting objects, mint capability destroyed, fair-launch auction on Sui, protocol-owned liquidity seeded, fee switch armed. The date moves if audits or legal review are not complete; the page is updated the same day.
4. 2027
   Burn engine and backstop go liveBuy-and-burn begins with the first settled mainnet contract. The staking backstop opens when the first insured capacity contracts settle. Exchange listings are not promised and never announced anywhere but the official-links page.

The date is a target, not a promise. Audits, legal review and testnet results decide it, and if any of them slip, this page changes before anything else does.

Risks and what is not decided

• Final share sizes may shift by a few points after audit and legal review; the hard cap, the burn floor and the no-mint rule will not.
• The exact funding split of the 6 % discount between broker commission and the reserve, and its caps, are published with the frozen parameters.
• The release-gating formula will be simulated against testnet volume before it is fixed in code.
• Whether early backers exist at all depends on the pre-launch round; if there is no round, that 12 % moves to the treasury under the same 36-month stream.
• Regulatory treatment differs by jurisdiction; the launch may exclude some regions.

Nothing on this page is an offer to sell, a solicitation, investment advice or a prediction of price. NMESH may never launch, and if it does it may be worth nothing. Read the risk disclosure and verify every address on the official links page before interacting with anything that claims to be NMESH.