PagerDuty SRE (Codex) - Evidence Grounding: leaderboard
Evidence Grounding (0-1), Codex fixed-harness cohort; 17 injected Kubernetes financial-app fault scenarios, one trial per agent, default config except model; same Grafana MELT, source code and alerts, no direct Kubernetes substrate or remediation. Three-provider judge ensemble. Higher is better.
Metric: Evidence Grounding (0-1), Codex fixed-harness cohort; 17 injected Kubernetes financial-app fault scenarios, one trial per agent, default config except model; same Grafana MELT, source code and alerts, no direct Kubernetes substrate or remediation. Three-provider judge ensemble. Higher is better.. Source: www.pagerduty.com. Saturation forecast: Around January 2027. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Terra | 0.9 | #42 |
| 2 | GPT-5.6 Sol | 0.85 | #18 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #7 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=pagerduty-sre-codex-evidence-grounding · How It Works · Data refreshed daily, snapshot 2026-10-16.