Claude Haiku 5.5: benchmark results
Provider: Anthropic. Access: API.
Unified ELO 1845 ± 18, rank #30 of 1632 rated models, from 133 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LiveBench Math Comp | 98.04 | Score | 98.5 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 84 | Omniscience Index | 97.9 |
| LLM Stats Score | 50.31 | LLM Stats Score (conservative rating) | 95.5 |
| AA-Briefcase - Analytical Quality Elo | 1907.63 | Elo | 94.6 |
| AA GDP.pdf - HR - Criterion Pass Rate | 84.12 | Criterion Pass Rate (%) | 94.3 |
| AA Long Context Reasoning | 82.67 | Accuracy (%) | 94.1 |
| Artificial Analysis Intelligence Index | 43.4 | Intelligence Index | 93.9 |
| AA Humanity's Last Exam | 44.39 | Accuracy (%) | 92.9 |
| LiveBench Theory of Mind | 86.54 | Score | 92.6 |
| AutomationBench-AA - Operations - Raw Objective Completion | 95.08 | Raw Objective Completion (%) | 92.4 |
| AA-Omniscience Index - Software Engineering (SWE) - JavaScript | 70 | Omniscience Index | 92.3 |
| AA-Omniscience Index - Software Engineering (SWE) - Swift | 64 | Omniscience Index | 92.3 |
Interactive version: theaggregate.ai/model?slug=claude-haiku-5-5 · How It Works · Data refreshed daily, snapshot 2026-10-08.