Model comparison
Command R vs Qwen2.5-Coder-32B
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 31.4 on the Noometry Index. Command R costs 2.8× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Cohere
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Command R scores higher in 1 category and Qwen2.5-Coder-32B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen2.5-Coder-32B leads 21.2 to 13.8.
- The biggest single-benchmark swing is LiveBench Coding: 17.9% for Command R and 56.9% for Qwen2.5-Coder-32B.
- Command R is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
- Command R accepts more context: 128K tokens versus 33K.
Side by side
| Command R | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | Cohere | Alibaba (Qwen) |
| Noometry Index | 31.4 | 33.4 |
| Released | 2024-08-30 | 2024-09-18 |
| Weights | Open | Open |
| Context window | 128K | 33K |
| Max output | 4K | 29K |
| Input $ / M tokens | $0.15 | $0.66 |
| Output $ / M tokens | $0.60 | $1 |
| Results tracked | 29 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Command R leads
Command R: 29.3 (#306), Qwen2.5-Coder-32B: 22.6 (#333)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| BigCodeBench Instruct | 37.1% | 49% |
| LiveBench Coding | 17.9% | 56.9% |
| LMArena Coding | 1169 | 1276 |
| BigCodeBench Complete | 45.2% | 58% |
| SWE-bench Verified (bash only) | — | 9% |
| Aider Polyglot | — | 16.4% |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Reasoning Qwen2.5-Coder-32B leads
Command R: 13.8 (#331), Qwen2.5-Coder-32B: 21.2 (#225)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Reasoning | 21.9% | 42.1% |
| LMArena Hard Prompts | 1164 | 1251 |
| LiveBench Data Analysis | 33.3% | 49.9% |
| LiveBench | 27.5% | 46.2% |
| DTBench | 46.4% | — |
| LMCA | 9.2% | — |
| Epoch Capabilities Index | — | 119.49 |
| HellaSwag | — | 83% |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
Command R: 28.0 (#246), Qwen2.5-Coder-32B: 33.3 (#204)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Math | 19.4% | 46.6% |
| LMArena Math | 1155 | 1251 |
| GSM8K | — | 93% |
Knowledge Qwen2.5-Coder-32B leads
Command R: 31.0 (#221), Qwen2.5-Coder-32B: 33.4 (#203)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1138 | 1221 |
| MMLU | 65.2% | 79.1% |
| ARC (AI2) Challenge | — | 70.5% |
Multilingual Qwen2.5-Coder-32B leads
Command R: 35.7 (#245), Qwen2.5-Coder-32B: 37.8 (#235)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1174 | 1205 |
| LMArena Chinese | 1182 | 1222 |
| LMArena Russian | 1174 | 1228 |
| LMArena French | 1162 | — |
| LMArena German | 1176 | — |
| LMArena Japanese | 1143 | — |
| LMArena Korean | 1163 | — |
| LMArena Spanish | 1151 | — |
Instruction Following Qwen2.5-Coder-32B leads
Command R: 58.1 (#261), Qwen2.5-Coder-32B: 61.4 (#245)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Instruction Following | 55.6% | 58.7% |
| LMArena Instruction Following | 1167 | 1223 |
Long Context Qwen2.5-Coder-32B leads
Command R: 36.3 (#231), Qwen2.5-Coder-32B: 38.0 (#208)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1198 | 1251 |
Writing & Preference Qwen2.5-Coder-32B leads
Command R: 38.2 (#254), Qwen2.5-Coder-32B: 41.6 (#240)
| Benchmark | Command R | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1187 | 1230 |
| LMArena Creative Writing | 1170 | 1174 |
| LMArena Multi-Turn | 1163 | 1222 |
| LiveBench Language | 16.7% | 23.3% |
Frequently asked questions
Is Command R better than Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 31.4 on the Noometry Index. Command R costs 2.8× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Which is cheaper, Command R or Qwen2.5-Coder-32B?
Command R is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is Command R or Qwen2.5-Coder-32B better for coding?
Command R scores higher on coding benchmarks: 29.3 versus 22.6 in the Noometry coding category.
Which has the bigger context window?
Command R does, with 128K tokens against 33K.
How many benchmarks do Command R and Qwen2.5-Coder-32B share?
22 benchmarks have published results for both models. Command R has 29 scored results on Noometry and Qwen2.5-Coder-32B has 31.