Benchmarks · agentic
ITSMBench
Uses live first-party pass@1 across 89 enterprise IT service-management tasks. Score: pass@1. Collected by Jake Boggs' benchmark dashboard.
6 models · 17 results · data as of 2026-10-29 · simulation.ressl.ai ↗
Leaderboard
Opus 5.5 max
All results
17 rows| # | Model | Effort | ITSMBench |
|---|---|---|---|
| 1 | Opus 5.5 | max | 56.8% |
| 2 | Opus 5 | max | 50.6% |
| 3 | Grok 4.5 | high | 45.4% |
| 4 | GPT-5.6 Sol | xhigh | 43.8% |
| 5 | GPT-5.6 Sol | high | 40.9% |
| 6 | Grok 4.5 | medium | 40.0% |
| 7 | GPT-5.6 Sol | medium | 31.2% |
| 8 | GPT-5.6 Terra | xhigh | 30.1% |
| 9 | Grok 4.5 | low | 29.7% |
| 10 | GPT-5.6 Terra | high | 29.0% |
| 11 | GPT-5.6 Sol | low | 24.0% |
| 12 | GPT-5.6 Terra | low | 20.2% |
| 13 | GPT-5.6 Terra | medium | 20.0% |
| 14 | GLM 5.2 | high | 19.8% |
| 15 | GLM 5.2 | xhigh | 19.1% |
| 16 | GPT-5.6 Sol | none | 17.3% |
| 17 | GPT-5.6 Terra | none | 11.2% |