Command beta (52.4B): benchmark results
Provider: Cohere.
Aggregate Rating 1344 ± 45, rank #7870 of 10916 rated models, from 32 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - CNN/DailyMail | 16.11 | ROUGE-2 (%) | 100 |
| HELM Classic - MS MARCO Regular | 47.21 | RR@10 (%) | 100 |
| HELM Classic - MS MARCO TREC | 76.23 | NDCG@10 (%) | 100 |
| HELM Classic - NaturalQuestions Open Book | 75.99 | F1 (%) | 98.5 |
| HELM Classic - NarrativeQA | 75.17 | F1 (%) | 95.4 |
| HELM | 87.45 | Mean win rate (imported) | 93.8 |
| HELM Classic - IMDB | 96.03 | Exact Match (%) | 92.4 |
| HELM Classic - HellaSwag | 81.1 | Exact Match (%) | 91.9 |
| RealToxicityPrompts | 0.01 | Toxic fraction | 91.5 |
| HELM Classic - LegalSupport | 60.6 | Exact Match (%) | 88.2 |
| HELM Classic - BoolQ | 85.63 | Exact Match (%) | 87.9 |
| HELM Classic - XSUM | 15.22 | ROUGE-2 (%) | 87.8 |
Interactive version: theaggregate.ai/model?slug=command-beta-52-4b · How It Works · Data refreshed daily, snapshot 2026-11-08.