> For the complete documentation index, see [llms.txt](https://docs.roboset.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.roboset.org/revenue-model/pay-per-work.md).

# Pay-per-work

### REWARDS AND THE POINTS ECONOMY

### Why Points

Direct per-episode cash pricing creates a liability the instant a submission is approved — the platform owes money before the data has earned anything. Points invert this. Contributors earn RP instantly on approval, while payouts only ever flow out of realized dataset revenue. The network can never owe more than it has earned.

Points also let quality multipliers, streak bonuses, and tier boosts compound without renegotiating cash prices, and they keep the reward loop alive even in periods with no dataset sales.

### RP as Revenue Shares

Every approved episode's RP is recorded and attached to the datasets that episode ships in. When a buyer licenses a dataset, the contributor allocation is distributed pro-rata by RP:

your payout = (your RP in the dataset ÷ total RP in the dataset) × contributor pool

A contributor holding 2% of the RP behind a dataset receives 2% of its contributor pool on every sale. RP is therefore not a fixed-value coupon — it is a weight on real revenue. Higher-value tasks carry more RP per episode, so the reward tiering survives intact, denominated in shares rather than fixed amounts.

### Quality Multipliers

At approval, the verification pipeline assigns a quality multiplier that scales the RP earned:

* Standard ×1.0
* High quality ×1.25
* Exceptional ×1.5

Additional stackable bonuses apply to specific contribution types (for example, failure-and-recovery recordings and non-dominant-hand variants), because that data is scarce and disproportionately valuable to buyers.

### The Redemption Floor

Pure revenue-share means contributors earn nothing until datasets sell, which discourages early supply. A periodic redemption pool provides a floor: contributors may redeem RP against a treasury-funded pool at a floating rate, capped at a published ceiling. Each RP converts exactly once — cash out now at the floor, or hold RP for the pro-rata share when the dataset sells, which is usually worth more.

### Rules That Keep RP Honest

* Earned, never bought. There is no purchase path for RP.
* Non-transferable. RP cannot move between wallets or accounts, which kills point-farming markets and Sybil laundering.
* Convert once. Each RP unit converts through either redemption or a dataset sale, never both.
* Bonuses mint RP, not cash. Gamification never creates unfunded liabilities.
* Minimum withdrawal. A minimum RP threshold per payout keeps processing sane.

Because RP is earned only through verified work and cannot be transferred, the ledger is naturally Sybil-resistant and forms a clean, contribution-weighted foundation for any future ownership distribution.

### Contributor Tiers

Consistency and quality raise a contributor's tier. Higher tiers unlock higher-paying tasks, premium bounties, and reward multipliers. Tier is earned through approved-episode volume and sustained quality scores, and it gates access to the most valuable work (for example, long-horizon sequence tasks unlock after a threshold of approved episodes).

<br>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.roboset.org/revenue-model/pay-per-work.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.