c4ai-command-r7B-12-2024 — benchmark results

Cohere's smallest Command R (7B) with a 128K context, open weights tuned for RAG, tool use, and agents. Provider: Cohere. Released 2024-12-11. Access: Open.

Unified ELO 1436 ± 22, rank #1065 of 1776 rated models, from 51 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard - IFEval77.13Score93.1
Open LLM Leaderboard - MATH Level 529.91Score81.8
EuroEval Spanish NLU - Sentiment Headlines ES47.95Sentiment classification Score (%)80.8
Arabic IFEval60.8Arabic Accuracy (%)79.5
EuroEval Spanish NLU50.76NLU Average Score (%)77.3
EuroEval Spanish NLU - MLQA ES64.05Reading comprehension Score (%)75.9
Open LLM Leaderboard - BBH36.02Score73.5
EuroEval Portuguese NLU - HAREM48.11Named entity recognition Score (%)71.7
EuroEval Italian NLU - MultiNERD IT69.8Named entity recognition Score (%)71.2
EuroEval Portuguese NLU - ScaLA PT20.49Linguistic acceptability Score (%)70.9
EuroEval Italian NLU - Sentipolc1658.28Sentiment classification Score (%)70.4
EuroEval Italian NLU54NLU Average Score (%)67.2

Interactive version: theaggregate.ai/model?slug=c4ai-command-r7b-12-2024 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.