Cohere · model passport
Command R
command-r- Open weights
- CC-BY-NC-4.0 licence
- Released 2024-03-11(curated)
- Overall · LMArena
- 1,227#313 of 396
- Best percentile
- 21.2Overall
- Cohort standing
- 26.1median 31.9
What the data says
Benchmarked on 7 of 27 tracked axes. Every figure below is the model's own published result on that board.
Weakest axes
- 10 pct — LiveBench (26.8%)
- 12 pct — MMLU-Pro (Artificial Analysis) (43.2%)
- 14.1 pct — LiveCodeBench (Artificial Analysis) (12.2%)
Uneven: 11.2 percentile points between its best and worst axis — 21.2 pct on Overall versus 10 pct on LiveBench.
- Overall21.2 pct1,227Verify →
- Instruction following19.9 pct1,200Verify →
- Hard prompts17.2 pct1,215Verify →
- Coding16.6 pct1,244Verify →
- LiveCodeBench (Artificial Analysis)14.1 pct12.2%Verify →
- MMLU-Pro (Artificial Analysis)12 pct43.2%Verify →
- LiveBench10 pct26.8%Verify →
- WebDev ArenaNot benchmarked
- Vision ArenaNot benchmarked
- Factuality ArenaNot benchmarked
- Search ArenaNot benchmarked
- Document ArenaNot benchmarked
- Agent ArenaNot benchmarked
- GPQA DiamondNot benchmarked
- AIME (math)Not benchmarked
- FrontierMathNot benchmarked
- SimpleQA VerifiedNot benchmarked
- ARC-AGI-2Not benchmarked
- Humanity's Last ExamNot benchmarked
- SimpleBenchNot benchmarked
- BFCL v4 (function calling)Not benchmarked
- SWE-bench Verified (live board)Not benchmarked
- Terminal-Bench 4.0Not benchmarked
- IFBench (Artificial Analysis)Not benchmarked
- HHEM Hallucination (Vectara)Not benchmarked
- SWE-bench VerifiedNot benchmarked
- Terminal-BenchNot benchmarked
Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.
command-r stands at 26.1 (raw mean 15.9) — measured on 7 of 27 axes.
command-r stands at 26.1 versus a cohort median of 31.9 across 104 models released within 180 days — below the peer median.
command-r scores in the 10th percentile on LiveBench.
command-r scores in the 12th percentile on MMLU-Pro (Artificial Analysis).
command-r ranks between #325 and #1,004 across 5 built-in weightings — best under "Conversation-first", worst under "All sources".
Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →