Claude 3 Opus (20240229): benchmark results
Claude 3 Opus (20240229) API snapshot, publicly available from 2024-03-04. The dated model ID identifies the snapshot; it is distinct from the public release date. Provider: Anthropic. Released 2024-03-04. Access: API.
Aggregate Rating 1608 ± 9, rank #412 of 1656 rated models, from 762 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM EWoK - Agent Properties | 94.46 | EM | 100 |
| HELM Finance - FinanceBench | 70 | Correct Answer | 100 |
| HELM Lite - WMT 2014 (ru-en) | 28.46 | BLEU-4 (%) | 100 |
| HELM MMLU - Security Studies | 88.57 | EM | 100 |
| Linguini | 24.05 | Exact Match (%; 0 in-context examples, 894 questions) | 100 |
| LogicKor - Grammar | 9.64 | Score (0-10) | 100 |
| HELM MMLU - College Biology | 97.22 | EM | 99.4 |
| Enkrypt AI - Safety Risk | 7.34 | Risk Score | 99.3 |
| HELM VHELM - Blink Perception - Object Localization | 100 | Prefix Quasi-Exact Match (%) | 98.9 |
| HELM AIR-Bench 2024 - #39-40.13: Discrimination in employment, benefits, or services - Mental Characteristics | 100 | Refusal Rate (%) | 98.8 |
| HELM AIR-Bench 2024 - #39-40.3: Discrimination in employment, benefits, or services - Color | 100 | Refusal Rate (%) | 98.8 |
| HELM AIR-Bench 2024 - #41-42.30: Unauthorized collection/gathering/stealing - Location data | 100 | Refusal Rate (%) | 98.8 |
Interactive version: theaggregate.ai/model?slug=claude-3-opus-20240229 · How It Works · Data refreshed daily, snapshot 2026-10-20.