Command beta (6.1B): benchmark results
Provider: Cohere.
Aggregate Rating 1252 ± 48, rank #9235 of 10916 rated models, from 32 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - NaturalQuestions Open Book | 71.75 | F1 (%) | 96.9 |
| HELM Classic - MS MARCO TREC | 70.94 | NDCG@10 (%) | 96.7 |
| HELM Classic - MS MARCO Regular | 43.35 | RR@10 (%) | 96.6 |
| HELM Classic - IMDB | 96.07 | Exact Match (%) | 93.9 |
| RealToxicityPrompts | 0.01 | Toxic fraction | 85.4 |
| HELM Classic - CNN/DailyMail | 15.31 | ROUGE-2 (%) | 82.9 |
| HELM Classic - Synthetic Reasoning Natural | 25.37 | F1 (%) | 79.4 |
| HELM Classic - NarrativeQA | 70.92 | F1 (%) | 72.3 |
| HELM Classic - BoolQ | 79.8 | Exact Match (%) | 71.2 |
| HELM Classic - LegalSupport | 56.65 | Exact Match (%) | 69.1 |
| HELM | 67.52 | Mean win rate (imported) | 67.7 |
| HELM Classic - QuAC | 37.51 | F1 (%) | 67.7 |
Interactive version: theaggregate.ai/model?slug=command-beta-6-1b · How It Works · Data refreshed daily, snapshot 2026-11-08.