Command R
Mar 2024
Command RCompare inon
Benchmarks of Command R16
BenchmarkByScoreRank
Writing15
Health & law1
No benchmark matches
—per task?
price not verifiedprice
—output speed?
128Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-5 Minihighwins 68%—$0.053
OpenGemma 4 31Bunstatedwins 78%—$0.10 / $0.34BetterGPT-6 Lunamaxwins 78%—$0.068Function calling only with none · Chat Completions+3BetterGPT-5.6 Lunaxhighwins 78%—$0.085BetterGPT-5.6 Terraxhighwins 79%—$0.63Prompts with >272K input tokensModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-5 Minihighwins 68%—$0.053GPT-6 Lunamaxwins 78%—$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 78%—$0.085GPT-5.6 Terraxhighwins 79%—$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 82%—$0.72Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
Gemma 4 31BunstatedOpenwins 78%—$0.10 / $0.34Gemma 4 26B A4BunstatedOpenwins 76%—$0.15 / $0.60
Qwen3.8 27BunstatedOpenwins 73%—$0.42 / $2.55max_tokens range · QwenCloud+4
Muse Glimmer 30BunstatedOpenwins 73%—$0.30 / $1.10Qwen 3.5 27BunstatedOpenwins 72%—$0.30 / $2.40ModelScoreCostPer taskWhat changes
Command AunstatedOpenwins 69%—$2.50 / $10Memory (4-bit): 21 GB → 67 GBCommand R+ (08-2024)unstatedOpenwins 60%—$2.50 / $10Memory (4-bit): not verifiedCommand R+Openwins 56%——Memory (4-bit): 21 GB → 62 GBCommand R (08-2024)unstatedOpensame—$0.15 / $0.60Memory (4-bit): not verifiedNorth Mini CodeOpenn/a——Command A+thinkingOpenn/a—$0.30 / $1.50Memory (4-bit): 21 GB → 131 GBTiny Aya GlobalOpenn/a——Context: 128K → 8K+1Command R7B (12-2024)Openn/a—$0.037 / $0.15Memory (4-bit): not verifiedAll Cohere models →Run it yourself
estimateFits on✗ 8 GB (phone, laptop)✗ 16 GB✓ 24 GB GPU · compressed✓ 48 GB GPU · 8-bit✓ 80 GB GPU · full quality
LicenceCC BY-NC 4.0no commercial use
History
2 versionsBenchmarks
all categoriesChat & Knowledge
Arena Text · overall1,227#179
Arena Text · Chinese1,146.4#166
Arena Text · English1,174.2#176
Arena Text · French1,162.4#141
Arena Text · German1,127.6#143
Arena Text · instruction following1,124.5#175
Arena Text · Japanese1,093.9#126
Arena Text · Korean1,091.7#133
Arena Text · non-English1,143.5#170
Arena Text · Polish1,207.5#128
Arena Text · Russian1,157.6#171
Arena Text · Spanish1,150.7#147
Vals LegalBench33#89
Arena Text · overall · full cohort1,163.1#173
Writing & Language
Arena Text · creative writing1,195#178
Arena Text · creative writing · full cohort1,154.1#170
FAQ
How much does Command R cost?›
There is no per-unit API price: price not verified.
How good is Command R, and what is it best at?›
#179 of 196 in Chat & Knowledge, #178 of 195 in Writing & Language.
Is Command R better than Llama 3 8B Instruct?›
On Arena Text · creative writing, Command R scores 1,195 vs 1,196 — about the same.
What is the context window of Command R?›
128K tokens.
When was Command R released? Is there a newer version?›
Released 2024-03-11. It is the current version in its line.
Is Command R open source?›
Yes, the weights are open (cc-by-nc-4.0 licence).
Can I run Command R locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Command R read images?›
No, it accepts text only.
What is Command R?›
Command R is an open-weights language model from Cohere, released Mar 2024.
What are the best alternatives to Command R?›
GPT-5 Mini (better), Gemma 4 31B (open weights).