Model comparison
Command R vs Deepseek Coder v2
Deepseek Coder v2 is the stronger model overall, scoring 35.3 to 31.0 on the Noometry Index.
Last verified . 19 shared benchmarks.
Cohere
DeepSeekSummary
- They share 19 benchmarks with published results for both. Command R scores higher in 0 categories and Deepseek Coder v2 in 8 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Deepseek Coder v2 leads 22.6 to 13.3.
- The biggest single-benchmark swing is BigCodeBench Complete: 45.2% for Command R and 59.7% for Deepseek Coder v2.
Side by side
| Command R | Deepseek Coder v2 | |
|---|---|---|
| Provider | Cohere | DeepSeek |
| Noometry Index | 31.0 | 35.3 |
| Released | 2024-08-30 | 2024-06-17 |
| Weights | Open | Open |
| Context window | 128K | — |
| Max output | 4K | — |
| Input $ / M tokens | $0.15 | — |
| Output $ / M tokens | $0.60 | — |
| Results tracked | 29 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Deepseek Coder v2 leads
Command R: 28.8 (#301), Deepseek Coder v2: 37.5 (#178)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| BigCodeBench Instruct | 37.1% | 48.2% |
| LMArena Coding | 1170 | 1251 |
| BigCodeBench Complete | 45.2% | 59.7% |
| LiveBench Coding | 17.9% | — |
| HumanEval+ | — | 82.3% |
| MBPP+ | — | 75.1% |
Reasoning Deepseek Coder v2 leads
Command R: 13.3 (#329), Deepseek Coder v2: 22.6 (#175)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Hard Prompts | 1164 | 1207 |
| LiveBench Reasoning | 21.9% | — |
| DTBench | 46.4% | — |
| LiveBench Data Analysis | 33.3% | — |
| LMCA | 9.2% | — |
| LiveBench | 27.5% | — |
| WinoGrande | — | 83.7% |
Math Deepseek Coder v2 leads
Command R: 25.8 (#244), Deepseek Coder v2: 32.3 (#188)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Math | 1155 | 1241 |
| LiveBench Math | 19.4% | — |
| GSM8K | — | 94.5% |
Knowledge Deepseek Coder v2 leads
Command R: 30.5 (#219), Deepseek Coder v2: 31.7 (#209)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Expert | 1138 | 1181 |
| ARC (AI2) Challenge | — | 64.3% |
| MMLU | 65.2% | — |
Multilingual Too close to call
Command R: 35.9 (#240), Deepseek Coder v2: 36.4 (#235)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Non-English | 1174 | 1182 |
| LMArena Chinese | 1182 | 1201 |
| LMArena French | 1161 | 1184 |
| LMArena German | 1176 | 1164 |
| LMArena Japanese | 1143 | 1126 |
| LMArena Korean | 1162 | 1104 |
| LMArena Russian | 1174 | 1188 |
| LMArena Spanish | 1150 | 1153 |
Instruction Following Deepseek Coder v2 leads
Command R: 58.3 (#256), Deepseek Coder v2: 61.9 (#237)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Instruction Following | 1167 | 1180 |
| LiveBench Instruction Following | 55.6% | — |
Long Context Too close to call
Command R: 36.3 (#226), Deepseek Coder v2: 37.0 (#219)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Longer Query | 1198 | 1219 |
Writing & Preference Too close to call
Command R: 38.4 (#248), Deepseek Coder v2: 38.4 (#249)
| Benchmark | Command R | Deepseek Coder v2 |
|---|---|---|
| LMArena Text | 1187 | 1191 |
| LMArena Creative Writing | 1170 | 1120 |
| LMArena Multi-Turn | 1163 | 1176 |
| LiveBench Language | 16.7% | — |
Frequently asked questions
Is Command R better than Deepseek Coder v2?
Deepseek Coder v2 is the stronger model overall, scoring 35.3 to 31.0 on the Noometry Index.
Is Command R or Deepseek Coder v2 better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 37.5 versus 28.8 in the Noometry coding category.
How many benchmarks do Command R and Deepseek Coder v2 share?
19 benchmarks have published results for both models. Command R has 29 scored results on Noometry and Deepseek Coder v2 has 24.