Model comparison
Command R vs GPT-4.1
GPT-4.1 is the stronger model overall, scoring 35.9 to 31.4 on the Noometry Index. Command R costs 13× less per token, which makes it the better buy when GPT-4.1's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
CohereSummary
- They share 19 benchmarks with published results for both. Command R scores higher in 2 categories and GPT-4.1 in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-4.1 leads 57.6 to 38.2.
- The biggest single-benchmark swing is DTBench: 46.4% for Command R and 68.3% for GPT-4.1.
- Command R is cheaper at $0.15 / $0.60 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 128K.
- Command R has downloadable open weights; the other is API-only.
Side by side
| Command R | GPT-4.1 | |
|---|---|---|
| Provider | Cohere | OpenAI |
| Noometry Index | 31.4 | 35.9 |
| Released | 2024-08-30 | 2025-04-14 |
| Weights | Open | Proprietary |
| Context window | 128K | 1.05M |
| Max output | 4K | 33K |
| Input $ / M tokens | $0.15 | $2 |
| Output $ / M tokens | $0.60 | $8 |
| Results tracked | 29 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4.1 leads
Command R: 29.3 (#306), GPT-4.1: 34.4 (#238)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Coding | 1169 | 1391 |
| SWE-bench Verified | — | 48.5% |
| SWE-bench Verified (bash only) | — | 39.6% |
| Aider Polyglot | — | 52.4% |
| WeirdML | — | 39% |
| BigCodeBench Instruct | 37.1% | — |
| LiveBench Coding | 17.9% | — |
| BigCodeBench Complete | 45.2% | — |
| CadEval | — | 42% |
| ALE-Bench | — | 558.1 |
Agentic & Tool Use Not comparable
Command R: —, GPT-4.1: 34.7 (#43)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 54% |
Reasoning Command R leads
Command R: 13.8 (#331), GPT-4.1: 11.7 (#339)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Hard Prompts | 1164 | 1384 |
| DTBench | 46.4% | 68.3% |
| LMCA | 9.2% | 25.6% |
| ARC-AGI-2 | — | 0.4% |
| SimpleBench | — | 27% |
| Kagi LLM Benchmark | — | 52.3% |
| ARC-AGI-1 | — | 5.5% |
| Chess Puzzles | — | 6% |
| EnigmaEval | — | 2.2% |
| LiveBench Reasoning | 21.9% | — |
| LiveBench Data Analysis | 33.3% | — |
| Epoch Capabilities Index | — | 136.78 |
| ForecastBench | — | 61.5 |
| LiveBench | 27.5% | — |
Math Command R leads
Command R: 28.0 (#246), GPT-4.1: 22.3 (#280)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Math | 1155 | 1370 |
| FrontierMath (Tiers 1-3) | — | 6% |
| OTIS Mock AIME 2024-2025 | — | 38.3% |
| Omni-MATH | — | 47.1% |
| LiveBench Math | 19.4% | — |
| MATH Level 5 | — | 83% |
| FrontierMath (Feb 2025 set) | — | 5.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge GPT-4.1 leads
Command R: 31.0 (#221), GPT-4.1: 37.1 (#160)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Expert | 1138 | 1364 |
| GPQA Diamond | — | 66.9% |
| Humanity's Last Exam | — | 5.4% |
| SimpleQA Verified | — | 31.1% |
| MMLU-Pro | — | 81.1% |
| Vectara Hallucination Rate | — | 5.6% |
| GPQA (HELM) | — | 65.9% |
| MMLU | 65.2% | — |
Multimodal Not comparable
Command R: —, GPT-4.1: 38.2 (#67)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Vision | — | 1211 |
| GeoBench | — | 72% |
Multilingual GPT-4.1 leads
Command R: 35.7 (#245), GPT-4.1: 49.4 (#133)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Non-English | 1174 | 1370 |
| LMArena Chinese | 1182 | 1382 |
| LMArena French | 1162 | 1382 |
| LMArena German | 1176 | 1381 |
| LMArena Japanese | 1143 | 1319 |
| LMArena Korean | 1163 | 1339 |
| LMArena Russian | 1174 | 1377 |
| LMArena Spanish | 1151 | 1376 |
Instruction Following GPT-4.1 leads
Command R: 58.1 (#261), GPT-4.1: 71.3 (#153)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Instruction Following | 1167 | 1367 |
| LiveBench Instruction Following | 55.6% | — |
| IFEval | — | 83.8% |
Long Context GPT-4.1 leads
Command R: 36.3 (#231), GPT-4.1: 40.0 (#163)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Longer Query | 1198 | 1385 |
| Fiction.LiveBench | — | 63.9% |
Writing & Preference GPT-4.1 leads
Command R: 38.2 (#254), GPT-4.1: 57.6 (#125)
| Benchmark | Command R | GPT-4.1 |
|---|---|---|
| LMArena Text | 1187 | 1383 |
| LMArena Creative Writing | 1170 | 1363 |
| LMArena Multi-Turn | 1163 | 1398 |
| EQ-Bench Creative Writing | — | 1420 |
| WildBench | — | 85.4% |
| LiveBench Language | 16.7% | — |
Frequently asked questions
Is Command R better than GPT-4.1?
GPT-4.1 is the stronger model overall, scoring 35.9 to 31.4 on the Noometry Index. Command R costs 13× less per token, which makes it the better buy when GPT-4.1's lead doesn't matter for your workload.
Which is cheaper, Command R or GPT-4.1?
Command R is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-4.1 lists at $2 and $8.
Is Command R or GPT-4.1 better for coding?
GPT-4.1 scores higher on coding benchmarks: 34.4 versus 29.3 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 128K.
How many benchmarks do Command R and GPT-4.1 share?
19 benchmarks have published results for both models. Command R has 29 scored results on Noometry and GPT-4.1 has 52.