Model comparison
Command R+ vs GPT-3.5-turbo
Command R+ is the stronger model overall, scoring 32.4 to 23.2 on the Noometry Index. GPT-3.5-turbo costs 5.8× less per token, which makes it the better buy when Command R+'s lead doesn't matter for your workload.
Last verified . 25 shared benchmarks.
CohereSummary
- They share 25 benchmarks with published results for both. Command R+ scores higher in 7 categories and GPT-3.5-turbo in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Command R+ leads 36.4 to 10.0.
- The biggest single-benchmark swing is BigCodeBench Complete: 41.9% for Command R+ and 50.6% for GPT-3.5-turbo.
- GPT-3.5-turbo is cheaper at $0.50 / $1.50 per million input/output tokens, against $2.50 / $10 for Command R+.
- Command R+ accepts more context: 128K tokens versus 16K.
- Command R+ has downloadable open weights; the other is API-only.
Side by side
| Command R+ | GPT-3.5-turbo | |
|---|---|---|
| Provider | Cohere | OpenAI |
| Noometry Index | 32.4 | 23.2 |
| Released | 2024-08-30 | 2023-03-01 |
| Weights | Open | Proprietary |
| Context window | 128K | 16K |
| Max output | 4K | 4K |
| Input $ / M tokens | $2.50 | $0.50 |
| Output $ / M tokens | $10 | $1.50 |
| Results tracked | 34 | 44 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Command R+ leads
Command R+: 29.1 (#309), GPT-3.5-turbo: 23.9 (#331)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| BigCodeBench Instruct | 33.8% | 39.1% |
| LMArena Coding | 1187 | 1136 |
| BigCodeBench Complete | 41.9% | 50.6% |
| HumanEval+ | 56.7% | 70.7% |
| MBPP+ | 63.5% | 69.7% |
| WeirdML | — | 3.5% |
| LiveBench Coding | 19.1% | — |
Agentic & Tool Use Not comparable
Command R+: —, GPT-3.5-turbo: —
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| METR Time Horizons | — | 21.5% |
Reasoning GPT-3.5-turbo leads
Command R+: 9.2 (#344), GPT-3.5-turbo: 13.8 (#332)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Hard Prompts | 1186 | 1108 |
| DTBench | 54.9% | 48.5% |
| LMCA | 5% | 9.7% |
| Epoch Capabilities Index | 119.34 | 118.55 |
| SimpleBench | 17.4% | — |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | 24.8% | — |
| Mystery Game Puzzles | — | 3% |
| LiveBench Data Analysis | 38.1% | — |
| Adversarial NLI | — | 58.1% |
| BIG-Bench Hard | — | 61.6% |
| CommonsenseQA 2.0 | — | 57% |
| ForecastBench | — | 50.4 |
| LiveBench | 31.8% | — |
| WinoGrande | — | 81.6% |
Math Command R+ leads
Command R+: 28.9 (#242), GPT-3.5-turbo: 6.3 (#327)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Math | 1188 | 1142 |
| FrontierMath (Tiers 1-3) | — | 0% |
| OTIS Mock AIME 2024-2025 | — | 2.2% |
| LiveBench Math | 21.3% | — |
| MATH Level 5 | — | 15.9% |
| GSM8K | — | 57.8% |
Knowledge Command R+ leads
Command R+: 36.4 (#169), GPT-3.5-turbo: 10.0 (#303)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Expert | 1174 | 1070 |
| MMLU | 69.4% | 71.4% |
| GPQA Diamond | — | 28% |
| Vectara Hallucination Rate | 6.9% | — |
| ARC (AI2) Challenge | — | 87.4% |
| BoolQ | — | 87% |
| OpenBookQA | — | 86% |
| TriviaQA | — | 85.8% |
Multilingual Command R+ leads
Command R+: 38.6 (#227), GPT-3.5-turbo: 31.5 (#258)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Non-English | 1216 | 1108 |
| LMArena Chinese | 1226 | 1075 |
| LMArena French | 1209 | 1118 |
| LMArena German | 1216 | 1090 |
| LMArena Japanese | 1166 | 1043 |
| LMArena Korean | 1138 | 1019 |
| LMArena Russian | 1227 | 1123 |
| LMArena Spanish | 1189 | 1121 |
Instruction Following Command R+ leads
Command R+: 60.0 (#254), GPT-3.5-turbo: 57.9 (#262)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Instruction Following | 1197 | 1119 |
| LiveBench Instruction Following | 57.6% | — |
Long Context Command R+ leads
Command R+: 37.3 (#219), GPT-3.5-turbo: 34.0 (#254)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Longer Query | 1230 | 1121 |
Writing & Preference Command R+ leads
Command R+: 43.5 (#228), GPT-3.5-turbo: 25.3 (#305)
| Benchmark | Command R+ | GPT-3.5-turbo |
|---|---|---|
| LMArena Text | 1229 | 1125 |
| LMArena Creative Writing | 1235 | 1092 |
| LMArena Multi-Turn | 1213 | 1117 |
| EQ-Bench Creative Writing | — | 451 |
| LiveBench Language | 29.7% | — |
Frequently asked questions
Is Command R+ better than GPT-3.5-turbo?
Command R+ is the stronger model overall, scoring 32.4 to 23.2 on the Noometry Index. GPT-3.5-turbo costs 5.8× less per token, which makes it the better buy when Command R+'s lead doesn't matter for your workload.
Which is cheaper, Command R+ or GPT-3.5-turbo?
GPT-3.5-turbo is cheaper. It lists at $0.50 per million input tokens and $1.50 per million output tokens; Command R+ lists at $2.50 and $10.
Is Command R+ or GPT-3.5-turbo better for coding?
Command R+ scores higher on coding benchmarks: 29.1 versus 23.9 in the Noometry coding category.
Which has the bigger context window?
Command R+ does, with 128K tokens against 16K.
How many benchmarks do Command R+ and GPT-3.5-turbo share?
25 benchmarks have published results for both models. Command R+ has 34 scored results on Noometry and GPT-3.5-turbo has 44.