Model comparison
Command R vs Qwen3.8 27B
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 31.4 on the Noometry Index. Command R costs 2.3× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
Cohere
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Command R scores higher in 0 categories and Qwen3.8 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.8 27B leads 65.8 to 38.2.
- The biggest single-benchmark swing is DTBench: 46.4% for Command R and 88% for Qwen3.8 27B.
- Command R is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.04 / $2.30 for Qwen3.8 27B.
- Qwen3.8 27B accepts more context: 262K tokens versus 128K.
Side by side
| Command R | Qwen3.8 27B | |
|---|---|---|
| Provider | Cohere | Alibaba (Qwen) |
| Noometry Index | 31.4 | 46.0 |
| Released | 2024-08-30 | 2026-08-14 |
| Weights | Open | Open |
| Context window | 128K | 262K |
| Max output | 4K | 33K |
| Input $ / M tokens | $0.15 | $0.04 |
| Output $ / M tokens | $0.60 | $2.30 |
| Results tracked | 29 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
Command R: 29.3 (#306), Qwen3.8 27B: 50.5 (#44)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Coding | 1169 | 1482 |
| LMArena WebDev | — | 1593 |
| SciCode | — | 46.6% |
| BigCodeBench Instruct | 37.1% | — |
| LiveBench Coding | 17.9% | — |
| BigCodeBench Complete | 45.2% | — |
Agentic & Tool Use Not comparable
Command R: —, Qwen3.8 27B: 32.9 (#57)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | — | 47.5% |
Reasoning Qwen3.8 27B leads
Command R: 13.8 (#331), Qwen3.8 27B: 41.0 (#54)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Hard Prompts | 1164 | 1460 |
| DTBench | 46.4% | 88% |
| LMCA | 9.2% | 41.4% |
| ARC-AGI-2 | — | 42.4% |
| NYT Connections (extended) | — | 54.5% |
| ARC-AGI-1 | — | 87.5% |
| CritPt | — | 5.4% |
| LiveBench Reasoning | 21.9% | — |
| LiveBench Data Analysis | 33.3% | — |
| Surface Evolver Bench | — | 45% |
| Epoch Capabilities Index | — | 149.38 |
| LiveBench | 27.5% | — |
Math Qwen3.8 27B leads
Command R: 28.0 (#246), Qwen3.8 27B: 37.1 (#161)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Math | 1155 | 1456 |
| ProofBench | — | 16% |
| LiveBench Math | 19.4% | — |
Knowledge Qwen3.8 27B leads
Command R: 31.0 (#221), Qwen3.8 27B: 41.6 (#109)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1138 | 1482 |
| MMLU | 65.2% | — |
Multimodal Not comparable
Command R: —, Qwen3.8 27B: 41.3 (#37)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | — | 1271 |
Multilingual Qwen3.8 27B leads
Command R: 35.7 (#245), Qwen3.8 27B: 53.7 (#60)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1174 | 1430 |
| LMArena Chinese | 1182 | 1504 |
| LMArena French | 1162 | 1465 |
| LMArena German | 1176 | 1438 |
| LMArena Japanese | 1143 | 1384 |
| LMArena Korean | 1163 | 1393 |
| LMArena Russian | 1174 | 1415 |
| LMArena Spanish | 1151 | 1448 |
Instruction Following Qwen3.8 27B leads
Command R: 58.1 (#261), Qwen3.8 27B: 75.8 (#53)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1167 | 1439 |
| LiveBench Instruction Following | 55.6% | — |
Long Context Qwen3.8 27B leads
Command R: 36.3 (#231), Qwen3.8 27B: 44.3 (#70)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1198 | 1450 |
Writing & Preference Qwen3.8 27B leads
Command R: 38.2 (#254), Qwen3.8 27B: 65.8 (#43)
| Benchmark | Command R | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1187 | 1441 |
| LMArena Creative Writing | 1170 | 1384 |
| LMArena Multi-Turn | 1163 | 1441 |
| EQ-Bench Creative Writing | — | 1671 |
| LiveBench Language | 16.7% | — |
Frequently asked questions
Is Command R better than Qwen3.8 27B?
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 31.4 on the Noometry Index. Command R costs 2.3× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.
Which is cheaper, Command R or Qwen3.8 27B?
Command R is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Qwen3.8 27B lists at $0.04 and $2.30.
Is Command R or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 29.3 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 27B does, with 262K tokens against 128K.
How many benchmarks do Command R and Qwen3.8 27B share?
19 benchmarks have published results for both models. Command R has 29 scored results on Noometry and Qwen3.8 27B has 31.