Incentive Lab

AI contributor reward calculator: worked examples

Incentive Lab separates customer revenue from token subsidy, then allocates a declared reward budget using accepted work and quality weights. Contributor caps and exact budget conservation keep the scenario inspectable. The worksheet does not establish Sybil resistance or distribute tokens.

Three fictional scenarios, calculated by the same engine as the tool. They illustrate behavior and failure cases, not customer outcomes or measured provider performance.

Rewards need a subsidy

Compare the balance before and after paying contributors.

Operating contribution, before rewards
5 USD
Including declared subsidy
105 USD
Unallocated reward budget
0 USD
Capped contribution allocation
ContributorAccepted tasksQuality factorEligibleReward
Contributor A100.9Yes18
Contributor B51Yes12
Contributor C301No0

Run this scenario Input JSON Calculated report

Inspect every input
{
  "currency": "USD",
  "customerRevenue": "80",
  "tokenSubsidy": "100",
  "computeCost": "50",
  "reviewCost": "20",
  "otherCost": "5",
  "budget": "30",
  "capPercent": 60,
  "contributors": [
    {
      "id": "Contributor A",
      "acceptedTasks": 10,
      "quality": 0.9,
      "independent": true
    },
    {
      "id": "Contributor B",
      "acceptedTasks": 5,
      "quality": 1,
      "independent": true
    },
    {
      "id": "Contributor C",
      "acceptedTasks": 30,
      "quality": 1,
      "independent": false
    }
  ]
}

Remove the token subsidy

Inspect whether customer revenue alone funds the reward plan.

Operating contribution, before rewards
5 USD
Including declared subsidy
5 USD
Unallocated reward budget
0 USD
Capped contribution allocation
ContributorAccepted tasksQuality factorEligibleReward
Contributor A100.9Yes18
Contributor B51Yes12
Contributor C301No0

Run this scenario Input JSON Calculated report

Inspect every input
{
  "currency": "USD",
  "customerRevenue": "80",
  "tokenSubsidy": "0",
  "computeCost": "50",
  "reviewCost": "20",
  "otherCost": "5",
  "budget": "30",
  "capPercent": 60,
  "contributors": [
    {
      "id": "Contributor A",
      "acceptedTasks": 10,
      "quality": 0.9,
      "independent": true
    },
    {
      "id": "Contributor B",
      "acceptedTasks": 5,
      "quality": 1,
      "independent": true
    },
    {
      "id": "Contributor C",
      "acceptedTasks": 30,
      "quality": 1,
      "independent": false
    }
  ]
}

Tighter contributor caps

A cap can leave part of the reward pool undistributed.

Operating contribution, before rewards
5 USD
Including declared subsidy
105 USD
Unallocated reward budget
15 USD
Capped contribution allocation
ContributorAccepted tasksQuality factorEligibleReward
Contributor A100.9Yes7.5
Contributor B51Yes7.5
Contributor C301No0

Run this scenario Input JSON Calculated report

Inspect every input
{
  "currency": "USD",
  "customerRevenue": "80",
  "tokenSubsidy": "100",
  "computeCost": "50",
  "reviewCost": "20",
  "otherCost": "5",
  "budget": "30",
  "capPercent": 25,
  "contributors": [
    {
      "id": "Contributor A",
      "acceptedTasks": 10,
      "quality": 0.9,
      "independent": true
    },
    {
      "id": "Contributor B",
      "acceptedTasks": 5,
      "quality": 1,
      "independent": true
    },
    {
      "id": "Contributor C",
      "acceptedTasks": 30,
      "quality": 1,
      "independent": false
    }
  ]
}

A mistake worth catching

A cap applies to a declared contributor ID, not a verified person. Splitting identities can defeat the economic intent unless a separate identity and review process is effective.

Use the method with your records

Use accepted tasks × declared quality as weights. Apply strict per-contributor caps with largest-remainder allocation; leave infeasible budget unallocated.

No live subnet data, fraud detection, Bittensor emission replication, staking recommendation or token distribution.

Read the complete method, sources and input contract. Updated 2026-09-19; by the AGI Scorecard team.

Continue your review

Agent EvidenceReview task-specific agent evaluations with sample deduplication, version boundaries and uncertainty intervals.Stable ReconcileReview invoices against your recorded stablecoin transfers. Keep partial payments, wrong assets and pending transactions visible.