Claude Haiku 4.5 (Non-reasoning) — benchmark results
Claude Haiku 4.5 evaluated with reasoning disabled. Provider: Anthropic. Released 2025-10-01. Access: API.
Unified ELO 1599 ± 30, rank #478 of 1776 rated models, from 41 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UGI - Writing | 52.16 | Writing Score | 90 |
| UGI - Natural Intelligence | 35.42 | NatInt Score | 84.3 |
| AA Omniscience | -7.67 | Score | 81.5 |
| AA Omniscience - Software Engineering (SWE) - Julia | 24 | Accuracy (%) | 70.4 |
| AA Terminal-Bench Hard | 27.27 | Accuracy (%) | 69.3 |
| AA MMLU-Pro | 80.01 | Accuracy (%) | 67.2 |
| Artificial Analysis Intelligence Index | 23.71 | Intelligence Index | 66.2 |
| AA LiveCodeBench | 51.11 | Pass@1 (%) | 57 |
| AA Omniscience - Software Engineering (SWE) - Swift | 40 | Accuracy (%) | 56 |
| WeirdML | 45.4 | Average Score | 54 |
| AA Long Context Reasoning | 43.67 | Accuracy (%) | 53.3 |
| AA SciCode | 34.38 | Accuracy (%) | 52.7 |
Interactive version: theaggregate.ai/model?slug=claude-haiku-4-5-non-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.