FET is a settlement token for agents, staking, and Alliance coordination

FET is trying to be a “work token” for an agent economy. That is the core bet. In the network documentation on native and ERC-20 FET, it is positioned as the utility token and medium of exchange for network services and transaction fees, and it is also the staking asset for Proof-of-Stake security.

The older token economics paper is explicit about the product linkage. FET is required for network exchanges, as a refundable registration mechanism, for staking, and as the value-transfer rail between autonomous economic agents in the token overview PDF.

Since mid-2024, “FET” is also the interim market ticker for the Artificial Superintelligence Alliance (ASI) token merger, where AGIX and OCEAN were consolidated into FET in Phase 1, with an eventual 1:1 migration of FET into ASI in Phase 2, as described in the Phase 1 update.

So the token has to clear two hurdles at once. It has to pay for real activity inside agent workflows. It also has to hold together a cross-project political economy where multiple communities are now monetarily entangled under one unit of account.

Supply reality: from “fixed cap” to merger-era expansion

Fetch’s original token economics described a fixed issuance of 1,152,997,575 tokens and stated that no further tokens would be created.

That “fixed cap” framing stopped being the right mental model in 2024. On May 3, 2024, Fetch.ai announced additional minting of 1,477,549,566 FET to support the ASI token exchange mechanism, bringing total supply to 2,630,547,141.

For a contrast case on how “cap narratives” can evolve over time, compare this with our ETC tokenomics review.

Post-merger, the cleanest way to reason about “supply” is to separate three layers:

(1) Genesis Fetch allocation and vesting rules that governed early stakeholder distribution.

(2) Merger mints that act like conversion inventory for other communities rather than “incentive emissions” in the classical sense.

(3) Chain-level monetary policy on the Fetch mainnet, which can introduce ongoing inflation if minting remains enabled.

Market data aggregators add a fourth layer: what is “live” and what is “locked” at any given moment. As of March 7, 2026, the supply data page reports Max Supply 2,714,384,546 FET and Circulating Supply 2,287,050,557 FET, and it also shows a contract address 0xaea46a60368a7bd060eec7df8cba43b7ef41ad85 for the ERC-20 representation.

That aggregator max-supply number does not match the May 3, 2024 disclosed post-mint supply of 2,630,547,141. This could reflect later alliance actions (such as the CUDOS integration) or differences in how aggregators reconcile multi-chain and migration-related supplies. The important point for analysts is practical: parameter stability is weaker when third-party “total supply” is not trivially reconcilable to a single, canonical issuer statement.

Emissions over time: vesting, migration inventory, and live inflation

The original Fetch token design was explicitly trying to avoid “forever inflation.” It put most supply questions into vesting and staged release schedules, rather than an open-ended mint rate. The token overview sets out vesting terms for private sale, founders, advisors, and the foundation, plus staged future releases and “mining” issuance.

Those vesting rules matter today because they show what the project thought was sustainable incentive pacing before the alliance-era expansion. In particular:

Founders and advisors were subject to a 3-month lockup after TGE, then quarterly vesting with 50% in year 1, 25% in year 2, and 25% in year 3.

Foundation tokens were described as linear vesting over 4 years.

Future releases were constrained to start no earlier than 12 months after TGE, to run up to 5 years, and not exceed one-third in any single year.

Mining issuance was described as introduced over 5 years, with release rate tied to network economic performance. That phrase is doing a lot of work. It is a design intent to link emissions to productivity, but the enforceability depends on who controls the release function and what “economic performance” means operationally.

The alliance merger changed the supply path with a single step-function. On May 3, 2024, Fetch.ai disclosed minting 866,700,367 FET to support the AGIX conversion and 610,849,199 FET to support the OCEAN conversion, as part of the overall 1,477,549,566 FET mint.

This is the key sustainability trade-off. Merger mints are not “ongoing inflation,” but they are still dilution relative to pre-merger holders, and they create a new unlock surface area that behaves like emissions when conversion inventory flows into liquid circulation.

The alliance then broadened again with CUDOS. The ASI governance proposal for integrating CUDOS states a conversion framework (including a 5% merge fee), a public vesting period of 3 months and treasury vesting of 10 months, and a stated supply increase of 88,946,755.672 FET.

Now the uncomfortable part for long-horizon modeling. Some Fetch documentation describes the network rewarding validators via block rewards and transaction fees, and it explicitly references an inflation rate (3% annual during the first three years of operation).

And independent chain parameter explorers currently report a minting inflation parameter of 3% for fetchhub-4.

I cannot reconcile, from public primary docs alone, whether that 3% is still actively expanding total supply on the current post-merger base, or whether it is a legacy parameter with bounded minting (for example, minting only until a cap, or minting from a pre-minted reserve). What I can say with confidence is that “FET is non-inflationary” is no longer a safe assumption for an analyst without verifying chain-level mint behavior and supply accounting end-to-end.

For a more overtly emissions-centered design to sanity-check your intuition, see our TAO tokenomics review.

Utility, fees, and where value can (and cannot) accrue

FET has three concrete sinks that matter for sustainability: (1) transaction fees, (2) staking and validator economics, (3) payment for network services inside the agent stack. Fetch’s network docs put fees and services front and center, and they also emphasize that FET exists as both ERC-20 and native mainnet representations, with the native token fueling the ecosystem.

On the validator side, Fetch has described validator revenue as coming from both block rewards and transaction fees, consistent with Cosmos-SDK style economics.

Fees are not just a revenue stream. They are an anti-spam mechanism. Fetch’s governance write-up references a concrete parameter push to raise minimum gas prices to 2 aFET via FIP-004, explicitly framed as a congestion and spam defense.

The token overview paper also sketches a design that decouples operational costs from the token itself, similar to Ethereum’s “gas,” and explicitly aims to increase stability of that operational fuel. This is a nuanced choice. It can reduce volatility in user costs, which is good for adoption, but it also weakens the reflexive “fee pressure drives token scarcity” story that many investors implicitly rely on.

On burn dynamics, the primary Fetch documents above do not specify a systematic burn policy for FET. That absence matters. If your long-run equilibrium depends on deflationary fee burns, you are modeling a mechanism that is not guaranteed by the canonical docs. The more realistic valuation route is throughput and demand for settlement, paired with a credible path to emissions tapering or productivity-linked emissions.

If you want a general checklist for mapping these sinks to value accrual, our design components guide is a useful companion.

The ASI merger adds another utility angle: consolidation. Phase 1 aimed to keep markets open under the FET ticker while AGIX and OCEAN were migrated, reducing “token fragmentation” across the combined stack.

Consolidation can improve monetary coordination. It can also blur accountability. When multiple foundations “remain independent” while sharing a token unit, cash flow transparency becomes more important, not less, because tokenholders are underwriting a wider surface of strategic spending.

Governance and parameter control

On paper, there are two governance layers. There is alliance governance, described as a governing council with named leadership across the founding entities.

Then there is chain governance on the Fetch mainnet, which uses a coin-voting model and a proposal framework (FIPs) that has already been used to adjust practical parameters like voting period, max gas, validator commission floors, and minimum gas prices.

One governance detail that matters for tokenomics is voter eligibility. Fetch’s March 31, 2025 upgrade note states that only those delegating FET on the Fetch.ai mainnet can vote, and holders on other representations like Ethereum, BSC, or Cardano are not affected by that governance process.

That creates a predictable political economy: validators and delegators on native mainnet have the strongest direct control over parameters that can affect fees, staking incentives, and (if enabled) inflation. If you are holding FET off-mainnet and not delegating, you are structurally exposed to decisions you do not directly vote on.

For emissions sustainability, this is a double-edged sword. Delegation-gated governance aligns decision-making with those actively underwriting security. It can also bias policy toward stakeholder groups that benefit from ongoing issuance, unless strong counterweights exist in the form of usage-driven fee revenue and explicit emissions sunsets.

For a broader view of best practices in parameter control and guardrails, see our methodology page.

Risk analysis: token-design pressure points

Allocations (documented)

Top 3 risks

  1. Supply-regime instability (dominant). Trigger: any further alliance expansions, treasury restructures, or policy changes that alter effective supply, unlock pace, or mint settings. Mechanism: step-function mints (as in May 3, 2024), plus potential ongoing inflation if minting remains enabled, can outpace real usage growth and keep the token in a dilution-overhang equilibrium. Who bears it: long-duration holders and builders holding working capital in FET, because their real claim on network demand is diluted unless demand grows faster than net issuance. Measurable indicators: disclosed mints and stated new totals (for example the jump to 2,630,547,141 on May 3, 2024), aggregator-reported supply and locked/unlocked deltas, and chain-level inflation parameters.

  2. Fee-market weakness. Trigger: agent activity and smart contract usage fail to grow enough to make fees meaningful relative to staking rewards. Mechanism: if validator economics remain dependent on emissions or block rewards rather than sustained fee revenue, the system inherits a structural pressure to keep inflation “on” for security, even if the token’s monetary premium depends on scarcity. Who bears it: all holders through long-run inflation pressure, and validators through unstable revenue if emissions are politically contested. Measurable indicators: governance moves focused on anti-spam fee floors (like minimum gas price changes), and the extent to which staking economics are framed around block rewards versus fees.

  3. Governance and representation mismatch across token venues. Trigger: material parameter changes (fees, inflation, validator set policy) executed through mainnet governance while a meaningful fraction of holders remain off-mainnet or un-delegated. Mechanism: voting power concentrates among delegators and validators on native mainnet, which can create policy outcomes that maximize staker yield or validator economics even when broader holder preferences would favor lower issuance and higher scarcity. Who bears it: passive holders and off-mainnet holders who cannot practically participate, plus application teams exposed to unpredictable fee policy. Measurable indicators: explicit statements that only delegators can vote, plus the cadence of governance proposals touching fees and validator economics.

Dominant risk: Supply-regime instability is the one that can break everything else, even if product-market fit arrives.

The reason is mechanical. If you want FET to be a settlement asset for an agent economy, then you need a credible long-run constraint on dilution relative to the rate of real economic throughput on the network. The 2024 ASI merger deliberately chose a step-function increase in supply to unify three monetary bases. That can be rational. It also means price performance will be dominated by whether demand grows faster than the new monetary base, and by how quickly conversion inventory becomes liquid.

Then CUDOS adds another dimension. The CUDOS proposal explicitly states a supply increase and vesting, which is better than pretending dilution does not exist. Still, it widens the number of stakeholders whose “fairness” expectations have to be satisfied by the same token. Coalition tokens drift toward political issuance when growth is slow. That is not moral failure. It is a predictable equilibrium.

Finally, emissions sustainability hinges on whether FET has a taper. Older Fetch materials talk like it should. The staking model explanation frames a transition to a fee-based incentive model, describing FET as non-inflationary in the long run.

But other Fetch materials explicitly describe inflation (3% annual for the first three years), and current parameter explorers still report a 3% inflation setting. That does not prove perpetual inflation in practice. It does prove that, from public docs alone, you cannot treat the monetary policy as “settled.”

If you care about long-horizon value accrual, the analytical job is to track three time series, not one: (1) effective circulating supply, (2) net issuance versus fee capture, (3) agent-driven transaction growth. Aggregator unlocked and locked figures are a pragmatic proxy for (1) if you are modeling market float today, but they should not be treated as canonical protocol truth.

One short advisory note: if you are doing tokenomics consulting work around coalition tokens like ASI/FET, treat “supply policy” as a governance product, not a static spreadsheet. If you need support designing those guardrails, see our tokenomics services page.

You will also want explicit guardrails for future expansions and a measurable linkage between any ongoing emissions and on-chain economic output; for examples of how we structure that kind of work, review our research reports.



This article is part of our Tokenomics Deep Dive series.