Grass is selling “decentralized data” while running a centralized control plane

Grass positions itself as a Sovereign Data Rollup that routes web requests through a globally distributed network, then checkpoints session proofs on-chain for provenance. That architecture has three operational actors that matter for tokenomics: nodes (end-user bandwidth), routers (regional relays), and a validator (the party initiating requests and verifying transactions).

Here’s the structural tension: the docs state the validator “initially” runs as a single, centralized entity, with a future plan to move to a validator committee. In decentralization terms, that means today’s network has a non-trivial trust anchor, and the token’s job is partly to fund and coordinate a transition that is not yet specified with hard governance thresholds.

On the user side, Grass Points function as the offchain accounting layer for contribution. Grass Tokens (GRASS) are the “reward” unit awarded via scheduled airdrop intervals that the Grass Foundation determines. That choice matters. It keeps the protocol flexible operationally, but it also concentrates discretion around distribution timing and any conversion logic between points and tokens.

Supply, allocations, and vesting: fixed supply, heavy insider tranches

GRASS supply is stated as fixed at 1,000,000,000. “Fixed” does not mean “no supply pressure.” It means all dilution is front-loaded into unlock schedules and discretionary distribution from pre-allocated pools.

CoinGecko’s current snapshot (as of March 5, 2026) shows 542,203,969 GRASS circulating, with 1,000,000,000 total supply. Treat that as a market-facing reality check: any tokenomics narrative that ignores unlock cadence is incomplete.

Grass’ initial allocation is described in its official docs as follows (percentages below are arithmetically derived from the fixed 1,000,000,000 total supply stated in the same document).

Airdrop One is the only distribution event with crisp, public mechanics in the docs. It allocated 100,000,000 GRASS (10% of total supply), split as 9% to Stage 1 point earners, 0.5% to GigaBuds NFT holders, and 0.5% to Desktop Node or Saga Application users. Claims opened on October 28, 2024.

The docs also state that the Airdrop One claiming window extension ended around March 27, 2025. History-wise, that date matters because it’s when a large chunk of “community allocation” stopped being claimable and became, in practice, administratively settled.

What the token does: stake-gated routing and reward emissions

GRASS has one clearly-live utility today: staking via delegation to routers. You stake by delegating tokens to routers, rewards are distributed “every second,” and unstaking has a 7-day unbonding period. Routers can charge a commission on rewards. This is a familiar lever in token economy design.

The mechanism design choice is explicit. Post-“decentralization,” the docs claim that routers with more delegated stake are awarded traffic more frequently, which increases rewards for router operators and their delegators. This is a stake-weighted scheduler. It is legible. It is also centralization-prone by default. For a comparable incentive pattern, see our SNX tokenomics review.

Grass’ router documentation goes further and formalizes reward math for delegates (including the router commission rate) and sketches a “post decentralization” reward structure tied to validated bandwidth, latency, and node reputation. Critical detail: the formula depends on coefficients that “need to be defined by the network.” Those coefficient choices are tokenomics. They decide who gets paid, and how quickly, for what kind of supply-side behavior.

Slashing is currently framed as a future possibility rather than an implemented constraint. The staking docs say there is no in-protocol implementation of slashing today, while acknowledging future updates may introduce it. Router docs still reference “potential slashing of staked assets” for failures like not maintaining a sufficiently accessible pool of nodes, with an uptime target stated as “greater than 75% (subject to change).”

As a decentralization purist, I read this as: the network wants cryptoeconomic discipline, but is not yet willing to lock itself into hard punishments. That is rational for early ops. It also means security is still mostly procedural and reputational, not credibly neutral.

Points, referrals, and the pre-token incentive engine

Grass Points are not just a marketing wrapper. They are the core distribution primitive that precedes token emissions. Points are split into Uptime Points and Network Points. The docs state that the updated points system became effective starting in Epoch 11 of Stage 2.

Network Points are particularly revealing because they are explicitly capped. Each day, Grass distributes a fixed pool of 1,000,000 Network Points across users according to their share of contributed bandwidth. If you believe future token distributions reference points, then this is your first “issuance schedule,” even if it is offchain.

The referral system is also aggressive. Uptime Points include ongoing referral commissions of 20% from primary referrals, 10% from secondary, and 5% from tertiary. That drives growth. It also creates a compounding distribution skew toward early networkers and large referrers, which later becomes token distribution skew if airdrops key off the points ledger.

Fees and fiscal flows: a formal fee market without a public value-capture map

Grass does publish a fee market model. Fees for web transactions are “dynamic,” and the docs provide a gas-like formula where cost depends on target geography, node reputation, traffic type, bandwidth consumed, and congestion, with coefficients that adjust over time. The traffic type dimension is concrete: partially-encrypted traffic (PET) gives the validator more ability to assess node quality, while fully-encrypted traffic (FET) provides stronger privacy but “will typically consume a higher gas rate.”

What’s missing from primary documentation is the market-facing fiscal flow diagram. The docs do not clearly specify, in one place, (1) what asset fees are paid in, (2) where those fees settle, and (3) what share goes to routers, nodes, delegators, validators, or treasury. The tokenomics page implies routers are subsidized early by the 30,000,000 Router Rewards pool “until the network matures and can sustain itself through network fees.” That is directionally useful, but it is not a parameterized revenue model. If you want more frameworks for thinking about value-capture mapping, browse our research reports.

A secondary disclosure from Kraken’s GRASS “white paper” (published July 17, 2025) states that on-chain governance is planned but not live, and that GRASS is planned to be accepted as payment for web-scraping transactions and dataset purchases when those features go live. I include this only to show how third parties are modeling expected utility. It is not a binding commitment by the issuer, and Kraken explicitly frames it as informational.

Governance and decentralization: the real parameters are still admin-controlled

Grass’ decentralization story, in its own docs, is a roadmap. The validator is described as initially centralized, later moving to a validator committee using a consensus mechanism. The validator section also admits the current validator “does not derive its security from any underlying L1,” and notes that downtime pushes users to submit session data directly on the L1 for settlement. A similar progressive-decentralization tension shows up elsewhere-see our ZK tokenomics review.

On the governance side, the tokenomics page says the Foundation & Ecosystem Growth allocation is “reserved for use as directed by DAO governance.” But the same documentation set places hard distribution timing under the Grass Foundation. The FAQ says tokens “will be rewarded via airdrop at intervals that are determined by the Grass Foundation.” That means the largest control surface affecting circulating supply is not credibly neutral today. For a concrete contrast in token-holder governance framing, see our ENS tokenomics review.

There is also a formal relationship disclosure: Wynd Labs is described as “a third party service provider” that provides the Grass Foundation with development and administrative services supporting the network. This is normal in practice. It also means “who holds the keys” is likely organizational, not protocol-level, until on-chain governance and a validator committee are actually live.

Grass also communicates a governance intent in its public-facing materials, framing GRASS as empowering holders to propose improvements and shape incentive mechanisms. Intent is not mechanism. In tokenomics, mechanism is threshold, quorum, timelock, and execution path. None of that is specified in primary docs today.

There is an official Grass Foundation forum endpoint, but it appears gated behind login, which limits public auditability of proposals and parameter debates. For a project whose value proposition leans on user ownership, that choice is a decentralization own-goal.

Risk register: tokenomics under a centralized chokepoint

The project’s token design can work economically. The problem is that the same design concentrates control in exactly the places that matter. The risks below are ranked by how directly they can break “user-owned infrastructure” as a practical claim.

Top 3 risks

  1. Control-plane centralization (dominant risk). Trigger: the single validator, foundation-controlled airdrop timing, or any admin-controlled coefficient update becomes contentious or captured. Mechanism: Grass’ own docs describe an initial single validator and multiple “to be defined by the network” coefficients in rewards and fees. Without on-chain governance live, those choices resolve to human governance and operational control, not token-holder sovereignty. Who bears it: delegators and airdrop recipients (valuation and reward-rate risk), router operators (policy risk), and buyers of GRASS (parameter stability risk). Measurable indicators: (i) concentration of validator role remaining at 1 entity over time, (ii) absence of publicly specified governance thresholds and execution controls, (iii) frequent or opaque parameter shifts in fee and reward coefficients, (iv) supply shocks driven by foundation-set distribution intervals.

    The hard part is not technical. It is political economy. A single validator can censor, reorder, or selectively validate traffic, and the network’s own security section acknowledges the validator does not currently inherit security from an L1. “Users can submit session data directly on the L1” is a fallback, not a baseline. It pushes operational burden onto participants exactly when the system is stressed.

    Tokenomics makes this worse because it ties economic rewards to the same control plane. If the validator is the verifier and the network defines the coefficients, then the entity that runs governance in practice can reshape who gets paid, where traffic is routed, and what “good behavior” means. In a stake-weighted router world, this invites cartel dynamics. If a small set of routers can attract stake and align with the validator operator, you get de facto permissioned routing with a token veneer.

    Progressive decentralization is a promise that expires if it is not enforced by deadlines and irreversible transitions. Grass’ docs describe the future committee, but do not publish a governance path with explicit thresholds and fail-safes. Until that exists, GRASS holders are underwriting an operational network that can remain structurally centralized while still paying out rewards.

  2. Stake-weighted router concentration. Trigger: a small number of routers accumulate delegated stake and become the default traffic recipients. Mechanism: the staking docs state that “post decentralization” more delegated stake leads to being awarded traffic more frequently, compounding rewards. Who bears it: small delegators (reward dilution and counterparty risk), smaller router operators (difficulty bootstrapping), and node operators (routing diversity decreases). Measurable indicators: (i) rising Gini coefficient of delegated stake across routers, (ii) increasing share of traffic routed through top-N routers, (iii) persistent commission rate convergence that signals weak competition.

  3. Value-capture opacity and parameter uncertainty. Trigger: market reprices GRASS because fee revenue pathways and treasury flows remain unclear or appear discretionary. Mechanism: the docs publish a fee model and talk about sustaining routers through network fees, but do not provide a consolidated, auditable model for how fees are charged, settled, and distributed. Who bears it: token holders (valuation fragility), router operators (business model uncertainty), and the foundation (credibility cost). Measurable indicators: (i) absence of public dashboards for protocol revenue split, (ii) continued reliance on token subsidies rather than fees, (iii) frequent changes in fee coefficients without governance ratification.

If you’re building something adjacent and want a second set of eyes, this is where tokenomics consulting is actually useful. The work is not slogans. It is pinning down governance thresholds, execution controls, and incentive parameters so decentralization is enforced by design, not by intent.



This article is part of our Tokenomics Deep Dive series.