Command Light: benchmark results

Provider: Cohere. Access: Open.

Unified ELO 1187 ± 52, rank #1536 of 1605 rated models, from 30 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
HELM Lite - LegalBench - International Citizenship Questions56.7Exact Match (%)58.9
HELM NaturalQuestions (Open)68.58F1 (%)42.2
HELM Lite - LegalBench - PROA87.37Exact Match (%)38.3
HELM Lite - MATH Level 1 - Number Theory16.67Equivalent (CoT) (%)19.4
HELM NarrativeQA62.95F1 (%)17.8
HELM Lite - MATH Level 1 - Intermediate Algebra11.54Equivalent (CoT) (%)16.1
HELM Lite - LegalBench - Corporate Lobbying17.35Exact Match (%)11.1
HELM Lite - MATH Level 1 - Algebra9.63Equivalent (CoT) (%)11.1
HELM Lite - MATH Level 19.81Equivalent (CoT) (%)8.9
HELM Lite - MATH Level 1 - Prealgebra11.63Equivalent (CoT) (%)8.9
HELM Lite - LegalBench39.67Exact Match (%)7.8
HELM Lite - MATH Level 1 - Precalculus14.04Equivalent (CoT) (%)7.8

Interactive version: theaggregate.ai/model?slug=command-light · How It Works · Data refreshed daily, snapshot 2026-09-26.