Command R (Mar '24) — benchmark results

Provider: Cohere. Released 2024-03-11. Access: Open.

Unified ELO 1171 ± 95, rank #1779 of 1841 rated models, from 7 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Humanity's Last Exam4.77Accuracy (%)28.1
Epoch AI - Scicode6.25Score6
AA GPQA Diamond28.38Accuracy (%)5.1
AA MMLU-Pro33.84Accuracy (%)4.4
Artificial Analysis Intelligence Index2.14Intelligence Index3.6
AA LiveCodeBench4.76Pass@1 (%)3.2
AA MATH-50016.4Accuracy (%)2.4

Interactive version: theaggregate.ai/model?slug=command-r-mar-24 · How It Works · Data refreshed daily, snapshot 2026-07-25.