12 / 19PROPOSEDA design; nothing here is deployed
TAO reserve proposal
ProposedPROPOSED. This chapter describes a design. Nothing in it is deployed, approved or running. Charging is off, TAON is not deployed, no TAO reserve exists and no audit has been done.
ARCHIVED PROPOSAL · NOT ADOPTED · TOKEN NOT YET JUSTIFIED
Two reserve models were considered, and they promise different things. Mixing them would make promises the contracts could not keep, so this record picks one for the first version.
Proposed · Not operating · Token not yet justified
Paid demand would guide the reserve
The Demand Reserve follows the subnets customers actually pay to use. Each paid request would leave a receipt naming the capability, route, provider, subnet, cost, latency and outcome; verified paid demand, not a vote or a price, would guide how eligible reserve capital is allocated across Bittensor.
Charging is off and revenue is $0, so the index is empty and the reserve holds zero. This is a design, not a balance. No token, pool, vault or reserve exists.
- Charging is OFF
- Customer revenue is $0
- TAON NOT DEPLOYED
- TAO reserve NOT DEPLOYED
Future reserve use is undecided. Potential uses require a published policy, custody design and measurable connection to product operation.
Read the proposalA static treasury holds assets whatever customers do. Here every allocation would originate from paid route evidence: no paid demand, no allocation.
The methodology, the caps and the empty indexWhat a reserve would hold, by accounting class
| Rule | Proposed |
|---|---|
| Asset | Native TAO on Bittensor. Bridged TAO on Robinhood Chain is not used until a first-party issuer document confirms its address, roles and redemption. |
| Custody | 3-of-5 multisig, five different people on separate hardware keys, at least two of them independent. No hot key and no automation can move it. |
| Acquisition rule | Once a month, after the monthly statement is signed, half of eligible surplus buys TAO once $250 is earmarked. Purchases above 1% over a three-quote reference price are cancelled. |
| Evidence | The address and signer list, the balance at each statement, every purchase and every outflow, and a signed monthly statement, including months with no purchase. |
| Outflows | Only paying a provider invoice for TAONIC service in TAO, or moving to a new multisig. No team payment, no token purchase, no staking, no lending. |
What the reserve is not:
- No user, buyer or holder has any claim on it, any redemption right or any share of it. Buying credit buys service only.
- It pays no return, and nothing is published about its growth.
- No staking rewards of any kind exist or may be described until a token passes all eight gates, audited contracts are deployed and a funded reserve exists with at least three published monthly records.
- A stranger can check the address and every transaction on chain. They cannot check that off-chain sales and costs are complete, because there is no external audit.
| Model A: Reserve growth | Model B: Staker entitlement | |
|---|---|---|
| User promise | TAO stays under protocol control and strengthens operations, routing incentives, integrations, liquidity or future governance. | TAO creates an enforceable, proportional entitlement for eligible TAON stakers. |
| Contract complexity | A controlled vault with bounded spending rules. | Staking, a reward index, per-holder accounting and a claim contract. |
| Solvency obligation | None to individuals: the reserve owes no one a share. | Every accrued entitlement must stay backed; a shortfall is a broken promise. |
| Accounting | Inflows, outflows and holdings, published. | All of that, plus every holder’s accrual, to the smallest unit, forever. |
| Regulatory sensitivity | Lower: no individual expectation of a share. | Higher: a pro-rata claim on an asset pool from holding a token. |
| Security | Custody, spending limits, a timelock on changes. | All of that, plus index arithmetic, claim reentrancy and partial-claim behaviour under shortfall. |
| Governance risk | Who approves spending, and how fast. | The same, plus any change to accrual rules affects everyone’s claim. |
| Audit scope | One vault and its controls. | Vault, staking, index and claims, and how they interact. |
PROPOSED Recorded 2026-09-23. For the first deployed version: Model A, Reserve growth.
Reserve growth creates fewer liabilities and is easier to verify: no per-holder solvency obligation, no claim contract, and a smaller audit scope. Version one has no staking, claims or holder yield. Any holder entitlement would need its own design, audit and governance decision, and none is proposed.
What this changes
- No surface describes TAON staking as earning TAO, and no surface shows an entitlement.
- Earlier research modelled a staker-entitlement reserve in detail. That work is kept as research, classified as a design for a possible later version, and shown nowhere in the product.
- What a reserve would be used for is undecided. Potential uses require a published policy, a custody design and a measurable connection to product operation, and each would be an allocation category published with its own accounting. Holding TAO is not claimed to improve routing, create demand, reward holders or earn yield.
If a reserve ever exists, every outflow would belong to a named category, approved in the open and behind a delay. A provider-growth allocation could be one, but only under an explicit written agreement that defines custody, ownership, accounting and withdrawal rights. No such allocation has been proposed to, or agreed with, any provider, and none exists. Provider costs are costs, never reserve contributions.
- Charged product revenue, or a live market fee, that produces eligible surplus.
- A written custody design and a governance authority with a timelock.
- A vault contract, fork-tested, and an independent audit.
- A public, reproducible accounting of every TAO in and out.
Applied to each revenue source separately, and published per period. The team and treasury share funds the work; the TAO reserve is a separate balance that is never drawn for pay.
ARCHIVED PROPOSAL · NOT ADOPTED · TOKEN NOT YET JUSTIFIED
| Use (PROPOSED) | Share |
|---|---|
| TAO reserve | 50% |
| Operations and provider costs | 15% |
| Team and treasury | 15% |
| Security and audits | 10% |
| Liquidity | 10% |
Conversion to TAO would be manual, two-person, reviewed weekly, with a minimum of $250; a week with no purchase is still published. The reserve asset would be the verified bridged TAO on Robinhood Chain, unless legal and technical review finds a better route.
Every state on this page matches Status. Where they ever differ, Status is right and this page is a defect.