Claude Haiku 5.5 (xHigh): benchmark results
Provider: Anthropic. Access: API.
Unified ELO 1820 ± 13, rank #107 of 2131 rated models, from 111 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA GDP.pdf - HR - Criterion Pass Rate | 83.22 | Criterion Pass Rate (%) | 92.8 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 74 | Omniscience Index | 92.6 |
| Artificial Analysis Intelligence Index | 41.25 | Intelligence Index | 92.4 |
| AA-Briefcase - Analytical Quality Elo | 1812.7 | Elo | 91.9 |
| AA CritPt | 22.57 | Accuracy (%) | 91.1 |
| AA Humanity's Last Exam | 42.68 | Accuracy (%) | 91 |
| AA-Omniscience Index - Software Engineering (SWE) - TypeScript | 63.33 | Omniscience Index | 90.7 |
| AA-Omniscience Index - Software Engineering (SWE) - JavaScript | 66.36 | Omniscience Index | 90.4 |
| AA-Omniscience Index - Software Engineering (SWE) - Python | 58.5 | Omniscience Index | 90.1 |
| AA-Omniscience Index - Software Engineering (SWE) - Julia | 36 | Omniscience Index | 88.2 |
| AA-Omniscience Index - Software Engineering (SWE) - Kotlin | 42 | Omniscience Index | 88.2 |
| AA-Briefcase - PDF - Rubric Pass Rate | 54.31 | Rubric Pass Rate (%) | 87.9 |
Interactive version: theaggregate.ai/model?slug=claude-haiku-5-5-xhigh · How It Works · Data refreshed daily, snapshot 2026-10-09.