Scoring Methodology
How we rank and rate x402 endpoints — transparent, objective, and fair.
🎯 Our Principles
Aligned with the x402 ethos of openness and fairness:
- Transparent — Every factor and weight is publicly documented
- Objective — Based only on measurable, verifiable data
- Fair — No pay-to-play, no hidden boosts, no manual overrides
- Open — Same rules for everyone, from solo devs to enterprises
- Permissionless — Automatic scoring, no application required
Quality Score (0-100)
Every endpoint receives a Quality Score from 0-100, calculated automatically from objective metrics:
Quality Score = (Uptime × 0.35) + (Speed × 0.25) + (Documentation × 0.20) + (Freshness × 0.20)
Uptime Score
35%
Speed Score
25%
Documentation Score
20%
Freshness Score
20%
Factor Breakdown
| Factor | How It's Measured | Score Range |
|---|---|---|
| Uptime | % of successful health checks over last 7 days. We ping endpoints every 30 minutes. | 0-100 (direct %) |
| Speed | Average response time. <500ms = 100, 500-1000ms = 75, 1000-2000ms = 50, 2000-5000ms = 25, >5000ms = 0 | 0-100 |
| Documentation | Has description (+40), has input schema (+30), has output schema (+30) | 0-100 |
| Freshness | Last verified active. Today = 100, 1-3 days = 80, 4-7 days = 60, 8-14 days = 40, 15-30 days = 20, >30 days = 0 | 0-100 |
Quality Tiers (A/B/C/D)
For quick reference, scores are grouped into tiers:
A
Score 75-100
Excellent reliability & documentation
B
Score 50-74
Good quality, minor issues
C
Score 25-49
Usable but needs improvement
D
Score 0-24
Unreliable or undocumented
Health Status
Separate from quality score, we track real-time health:
| Status | Definition |
|---|---|
| 🟢 Healthy | Responded successfully to last 3+ health checks, response time under 2 seconds |
| 🟡 Degraded | Responding but slow (>2s) or intermittent failures in last 24 hours |
| 🔴 Unhealthy | Failed last 3+ health checks or completely unresponsive |
| ⚪ Unknown | Not yet checked or no recent health data |
What We Don't Do
No paid rankings. You cannot pay to boost your score or ranking.
No manual overrides. Scores are calculated automatically — we don't manually adjust them.
No hidden factors. Every input to the score is documented above.
No favoritism. Same algorithm for everyone, including our own endpoints.
No manual overrides. Scores are calculated automatically — we don't manually adjust them.
No hidden factors. Every input to the score is documented above.
No favoritism. Same algorithm for everyone, including our own endpoints.
Improving Your Score
Want a better score? Here's what actually helps:
- Keep your endpoint online and responsive (affects Uptime & Speed)
- Add detailed descriptions to your x402 config (affects Documentation)
- Include input/output schemas in your Bazaar extensions (affects Documentation)
- Ensure your endpoint returns proper 402 responses (affects Health)
Data Sources
We index endpoints from:
- Coinbase Bazaar — Official x402 discovery API
- Manual submissions — Via our submit form
- Ecosystem scanning — We crawl known x402 deployments
Questions?
Our scoring is designed to be fair and transparent. If you have questions or find an issue, reach out on Twitter @theagentindex.