Child of SWE-bench Verified
Metadata
| Source Table | benchmark_results |
| Source ID | 9iqsCKcMgg |
| Parent | SWE-bench Verified |
| Children | — |
| Created | Apr 24, 2026, 6:42 PM |
| Updated | Apr 24, 2026, 6:42 PM |
| Synced | Apr 24, 2026, 6:42 PM |
Record Data
id | 9iqsCKcMgg |
benchmarkId | WOSlsBTTmV |
modelId | Claude Haiku 4.5(ai-model) |
score | 73.3 |
unit | percent |
date | 2025-10-15 |
sourceUrl | — |
notes | Score reported by Anthropic averaged over 50 runs. One of the highest scores in its price tier. Tested via Anthropic's agent scaffold. |
testedBy | unknown |
testedByOrgId | — |
evaluationDate | — |
methodologyNotes | — |
Source Check Verdicts
confirmed98% confidence
Last checked: 4/24/2026
Inline sourcing: confirmed
Debug info
Thing ID: 9iqsCKcMgg
Source Table: benchmark_results
Source ID: 9iqsCKcMgg
Parent Thing ID: WOSlsBTTmV