Claude Haiku 4.5 (20251001) (Thinking) — benchmark results
Claude Haiku 4.5 (20251001) evaluated with thinking enabled. Provider: Anthropic. Released 2025-10-01. Access: API.
Unified ELO 1654 ± 17, rank #358 of 1776 rated models, from 32 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UGI - Writing | 49.08 | Writing Score | 88.5 |
| UGI - Natural Intelligence | 36.76 | NatInt Score | 85.8 |
| Vals AI MedScribe | 85.23 | Accuracy (%) | 81.2 |
| Vals AI MGSM | 92.15 | Accuracy (%) | 66.4 |
| Vals AI AIME | 82.71 | Accuracy (%) | 54.7 |
| Vals AI CorpFin v2 | 60.61 | Accuracy (%) | 53.3 |
| Vals AI LegalBench | 81.24 | Accuracy (%) | 51.2 |
| Vals AI MortgageTax | 62.16 | Accuracy (%) | 48.8 |
| Vals AI Terminal-Bench 2.0 | 38.2 | Accuracy (%) | 46.2 |
| Vals AI CaseLaw v2 | 56.48 | Accuracy (%) | 45.4 |
| Vals AI GPQA | 72.22 | Accuracy (%) | 38.9 |
| Vals AI Finance Agent | 46.93 | Accuracy (%) | 38 |
Interactive version: theaggregate.ai/model?slug=claude-haiku-4-5-20251001-thinking · How the rankings work · Data refreshed daily, snapshot 2026-07-22.