The $BV7X Token Model

Most token models try to do too much. They mint a single asset and ask it to be gas, governance, yield, security, and brand all at once. The result is a token that is structurally compromised at every transaction point — nobody is sure whether they hold it, stake it, spend it, or vote with it.

$BV7X is built around a simpler claim: workers earn token, capital earns yield, treasury sustains the protocol. Three flows. One mental model. Everything else is implementation detail.

Three actor types, three reward streams

The network has three groups of participants, and each one earns from a different stream:

• Agents — the forecasters. They commit predictions, settle them on prediction markets, and earn $BV7X emissions weighted by accuracy.
• Integrators — the exchanges, trading apps, and prediction-market venues that route flow to BV-7X. They earn $BV7X emissions pro-rata to the fee revenue they bring to the protocol.
• Stakers — anyone who locks $BV7X. They earn 60% of all licensing revenue, in the currency it was paid in — USDC stays USDC, $BV7X stays $BV7X — distributed weekly.

The remaining 40% of licensing revenue accrues to a multisig-controlled protocol treasury. Audits, oracle reliability, integrator BD, contingency — funded by the network itself, in perpetuity, in stablecoins. No ongoing dilution required.

Flow 1 — Pay to mine (the burn)

Every miner-side $BV7X outflow burns. There are three sources:

• Submission gas. Each prediction commit costs 10–50 $BV7X. Forces continuous miner-side acquisition; not a one-time stake.
• Bilateral wager fees. When agents wager $BV7X directly against each other on the same event, the winner takes 98% of the pot and the protocol burns 2%.
• Slashing. Non-reveal, sustained miss rate above 70%, or detected collusion sends 100% of the slashed stake to the burn address.

That is the entire deflation surface. It scales with active forecasting, not with consumer demand. The more agents work, the more $BV7X leaves circulation. The 1B supply cap is hard.

Flow 2 — Pay to consume (the licensing revenue)

Consumers and integrators pay for intelligence. They can settle in either $BV7X or USDC; native $BV7X settlement carries a 3% discount.

Revenue source	Currency	Destination
Intelligence access (per-call + subscription)	$BV7X / USDC	60% stakers · 40% treasury
Integrator performance fee (20% of net P&L)	USDC	60% stakers · 40% treasury
Routing fees (Lighter 5 bps, Polymarket)	USDC	60% stakers · 40% treasury
Submission gas (10–50 $BV7X / commit)	$BV7X	100% burned
Bilateral wager fee (2% of pot)	$BV7X	100% burned
Slashed stake	$BV7X	100% burned

Two destinations only: stakers, or burn. There is no buyback router, no carve-out for the founders, no skim of the staker yield surface. The treasury covers protocol operations from a clean revenue split, not from an ongoing tax on holders.

Flow 3 — Mine to earn ($BV7X emissions)

The token model rewards the people who do the work. Three groups receive $BV7X emissions:

• Agents (40% of supply, 96 months, demand-throttled) — weighted by stake × accuracy². Quadratic accuracy weighting concentrates emissions on real edge and penalizes wash-mining.
• Integrators (5% of supply, 24 months linear) — pro-rata to the fee revenue each integrator routes to the protocol. Bootstrap incentive that aligns Vibe Trading and the integrators that follow with $BV7X price, not just their commercial fee.
• Stakers (5% of supply, 24 months linear) — an xBV7X bootstrap stream that seeds early staking yield while licensing revenue ramps.

Mining emissions are demand-throttled: if revenue undershoots, the daily rate slows; if usage outpaces supply, it accelerates. The protocol cannot dilute holders faster than the network grows.

Bilateral wagering

Two agents post opposing predictions on the same event and each lock equal $BV7X stake into a vault contract. On resolution, the winner takes the pot. The protocol takes 2% and burns it.

This adds a layer of accuracy that does not depend on the external prediction market. It surfaces three things at once:

• Conviction. Agents who refuse to wager are signaling a weak forecast.
• Pairwise reputation. Head-to-head records become a richer signal than aggregate accuracy alone — the credit-score equivalent of beating specific competitors at specific markets.
• Anti-sybil at the protocol layer. Wash-mining your own farm of agents is now economically negative. You cannot game your own counterparty.

Wagering is optional. Unmatched predictions fall back to the standard mining flow. Wager outcomes emit on-chain and feed the ERC-8004 reputation registry the same way any other settlement does.

Allocation

Total supply is 1,000,000,000 $BV7X — fixed, with a hard cap.

Allocation	%	Cliff	Vesting
Mining emissions (agents)	40%	0	96mo demand-throttled
Integrator rewards	5%	0	24mo linear
Staker bootstrap	5%	0	24mo linear
Ecosystem / Treasury	15%	0	36mo
Migration	15%	0	12mo
Team & Advisors	10%	12mo	36mo
Investors	8%	6mo	24mo
Liquidity & MM	2%	0	0 (TGE)
Public Sale	0.5%	0	0 (TGE)

The token starts with a tight float (~4.75% at TGE) and unlocks gradually. Cumulative checkpoints, excluding the throttled mining stream: TGE 4.75% · +1 month 6.4% · +12 months 22%.

Why the structure works

Most token models confuse two questions that should stay separate: who does the work, and who provides the capital. They reward both with the same instrument and end up rewarding neither well.

$BV7X separates them. Agents and integrators — the workers — earn $BV7X emissions, which gives them a long-term stake in the network they're contributing to. Stakers — the capital — earn yield in the currency the protocol earned it in, weekly. The treasury keeps a piece of every revenue line so the network can fund its own audits and partnerships without a cap-table tax.

Burn is dedicated to the deflationary lever it is best at: penalizing miner waste, charging gas, and slashing misbehavior. It does not have to do double duty as the staker yield engine, because staker yield comes from a real revenue stream.

The category $BV7X belongs to is decentralized-compute network tokens — RENDER, IO, FIL, AKT, TAO. Each is a metered compute network with a native gas token. BV-7X applies the same architecture to predictive AI compute, with bilateral wagering as the additional accuracy lever and an xBV7X real-yield surface modeled on what GMX and dYdX v4 proved out for trading-fee distribution.

---

The infrastructure that makes the token work is already in motion. The arena is live. The forecast is live. Settlement on Polymarket is live. The on-chain attestation pipeline is live. The contracts to enforce this token model — FeeDistributor, WagerVault, Integrator-Rewards emissions controller — ship in Phase 1 of the network upgrade, ahead of the third-party audit.

What changes at TGE is not the product. It is the way the network pays the people who built it.