Who's actually on the cost/accuracy frontier?Design-stage
The Leaderboard is Fugal's endgame: a public, continuously-re-measured board where submitted routing systems are scored each epoch on fresh, burn-once tasks by a deterministic grader. Because model prices and skills drift, the board is meant to visibly decay and refresh — a live record of the frontier, not a frozen screenshot. It is not live yet; this page is the shape of what's coming.
| Rank | System | Accuracy | $/query | Utility |
|---|---|---|---|---|
| 1 | fugal-router | 0.9xx | $0.00xx | ····· |
| 2 | ░░░░░░░░░ | 0.9xx | $0.00xx | ····· |
| 3 | ░░░░░░░░░ | 0.9xx | $0.0xxx | ····· |
| 4 | ░░░░░░░░░ | 0.8xx | $0.0xxx | ····· |
Burn-once tasks
Each epoch scores on freshly generated, un-memorized questions — solve it once, it's retired. No teaching to the test.
Deterministic grading
A reference validator grades every submission the same way, re-gradeable from the archived artifact — the frontier is auditable, not asserted.
Ranked by utility
Systems are ranked on utility = accuracy − λ·cost, so a cheaper system that stays correct beats a pricier one that merely ties.
How it connects to the demo
Fugal ships as the open, fork-me reference system that seeds row 1. The same refresh loop behind the Research scoreboard — re-price, re-score, recalibrate, publish a dated snapshot — is what the Leaderboard will run on a cadence. Read the full design on the About page.