TorchRank

What is TAO?

TAO is Bittensor’s native token. It secures validator behavior via staking, pays contributors for useful AI work across subnets, and aligns incentives so the network favors reliability over hype.

Utility
Emissions
Staking & Delegation
Validator Economics

TL;DR

Role

Coordination token

TAO governs staking/delegation for validators and pays providers across subnets for measured usefulness.

Who uses it

Delegators, Validators, Providers

Delegators back validators. Validators curate quality. Providers supply models/compute to subnets.

Why now

Open AI markets

Demand is high; TAO routes incentives to consistent performance rather than marketing claims.

What TAO does

Staking & Security

Validators stake TAO and run scoring logic. Accurate, consistent scoring attracts delegation and rewards.

Signal: uptime, scoring stability, and delegation growth over time.

Incentives

Providers (miners/models) earn TAO according to validator-measured usefulness on each subnet.

Better outputs → higher scores → more emissions share.

Access & Fees

Some subnets may require TAO (or TAO-denominated fees) for participation, priority, or bandwidth.

Governance (parameters)

Holders influence parameters that shape incentives and safety rails. Details vary by upgrade epoch.

Emissions & distribution

Network emissions fund rewards to subnets. Distribution can tilt toward subnets that deliver measurable value.

High-level flow

  • Total emissions are split across subnets.
  • Within a subnet, validators direct rewards to providers based on scored usefulness.
  • Validators and their delegators share validator-side rewards per fee schedule.

Plain-English math

// emissions (simplified)
subnet_reward   = total_emissions * subnet_weight
provider_reward = subnet_reward * provider_score_share
validator_take  = subnet_reward * validator_share
delegator_take  = validator_take * (1 - validator_fee)

Exact parameters change over time; focus on consistency of your share, not one-day spikes.

Operator takeaway: track 7–30 day trends. The goal is steady share, not lottery days.

Staking & delegation (how-to)

1) Secure a wallet
Generate and back up keys offline. Store seed safely; use a hardware wallet if supported.
2) Acquire TAO
Move TAO to your staking wallet address. Double-check chain/network before sending.
3) Shortlist validators
Use TorchRank’s leaderboard to filter by consistency, fee, uptime, and recent delegation flows.
4) Delegate
Submit a delegation tx to your chosen validator(s). Start small; verify rewards accrue as expected.
5) Monitor & rebalance
Set a rule: rotate or reduce if 7-day scoring/uptime drops or fees jump.
6) Document settings
Keep a simple log (validator, fee, start date, reason). Helps avoid bias and track decisions.
Open Leaderboard Step-by-step staking guide

Validator economics

Revenue side

  • Share of subnet emissions flowing to validators.
  • Delegation base and fee schedule.
  • Scoring accuracy/consistency (drives future delegation).
// very simplified
rev_tao = subnet_emissions * your_validator_share
payout  = rev_tao * (1 - fee_to_operator)  // delegators get the complement

Cost side

  • Infra (CPU/GPU as needed), bandwidth, storage, monitoring.
  • Ops time (alerts, upgrades, scoring logic upkeep).
  • Risk buffers (slashing/penalties if applicable).
net_yield = payout - (infra_cost + ops_cost + risk_buffer)

Scale only when 14–30 day net yield is stable and delegation is growing.

Risks & safeguards

Common risks

  • Performance decay → lower rewards / delegation outflows.
  • Key mismanagement → loss of funds or control.
  • Downtime / missed updates → penalties or reputation loss.
  • Fee changes → net yield compression for delegators.

Safeguards

  • Hardware wallet + offline backups; rotate hot keys.
  • Monitoring + paging; target SLOs for uptime & latency.
  • Change control: test scoring logic before production.
  • Guardrails: auto-rebalance delegation on 7-day dips.
Reminder: publish a light “validator README” so delegators know your SLOs and update cadence.

Choosing a validator

Signal

Consistency

Stable scoring & uptime over 7–30 days > one-day peaks. Look for smooth lines, not spikes.

Economics

Fee & flow

Fee should be justified by performance. Track delegation inflow/outflow around fee changes.

Ops

Transparency

Clear notes, changelogs, and incident handling earn trust and stickier delegation.

Use TorchRank filters Roles overview

FAQ

Is TAO only for validators?

No. Delegators use TAO to back validators, and providers earn TAO for useful work on subnets.

How often do rewards update?

Continuously at the network level; track 7–30 day windows to avoid noise.

What if my validator underperforms?

Reduce or rotate per your rulebook (e.g., 7-day consistency threshold). Document each move.

Can fees change?

Yes. Monitor fee changes versus performance—chasing low fees without quality can reduce yield.

← What is Bittensor? How to run a validator →