Claude v1.3: benchmark results
Provider: Anthropic. Access: API.
Unified ELO 1487 ± 34, rank #1198 of 2656 rated models, from 31 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Lite - WMT 2014 (ru-en) | 28.03 | BLEU-4 (%) | 94.4 |
| HELM WMT 2014 | 21.85 | BLEU-4 (%) | 80 |
| HELM v2 Lite - MedQA | 75 | Exact Match (%) | 80 |
| HELM Lite - WMT 2014 (cs-en) | 23.1 | BLEU-4 (%) | 78.9 |
| HELM Lite - WMT 2014 (de-en) | 20.21 | BLEU-4 (%) | 73.3 |
| HELM Lite - WMT 2014 (fr-en) | 22.75 | BLEU-4 (%) | 70.8 |
| HELM Lite - WMT 2014 (hi-en) | 15.17 | BLEU-4 (%) | 66.7 |
| HELM NaturalQuestions (Closed) | 40.89 | F1 (%) | 66.7 |
| HELM Lite - LegalBench - Function of Decision Section | 41.69 | Exact Match (%) | 65.6 |
| HELM Lite | 57.19 | Mean win rate (self-reported) | 64.5 |
| HELM Lite - MATH Level 1 - Prealgebra | 82.56 | Equivalent (CoT) (%) | 63.9 |
| HELM Lite - LegalBench - Abercrombie | 65.26 | Exact Match (%) | 61.1 |
Interactive version: theaggregate.ai/model?slug=claude-v1-3 · How It Works · Data refreshed daily, snapshot 2026-09-19.