Model comparison
Command A vs GPT-5 Nano
Command A is the stronger model overall, scoring 36.5 to 33.5 on the Noometry Index. GPT-5 Nano costs 32× less per token, which makes it the better buy when Command A's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
CohereSummary
- They share 22 benchmarks with published results for both. Command A scores higher in 7 categories and GPT-5 Nano in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Command A leads 35.9 to 25.8.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 28.8% for Command A and 62.2% for GPT-5 Nano.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $2.50 / $10 for Command A.
- GPT-5 Nano accepts more context: 400K tokens versus 256K.
- Command A has downloadable open weights; the other is API-only.
Side by side
| Command A | GPT-5 Nano | |
|---|---|---|
| Provider | Cohere | OpenAI |
| Noometry Index | 36.5 | 33.5 |
| Released | 2025-03-13 | 2025-08-07 |
| Weights | Open | Proprietary |
| Context window | 256K | 400K |
| Max output | 8K | 128K |
| Input $ / M tokens | $2.50 | $0.05 |
| Output $ / M tokens | $10 | $0.40 |
| Results tracked | 24 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Nano leads
Command A: 27.2 (#322), GPT-5 Nano: 33.6 (#254)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Coding | 1330 | 1351 |
| SWE-bench Verified (bash only) | — | 34.8% |
| Aider Polyglot | 12% | — |
| WeirdML | — | 38.1% |
| ALE-Bench | — | 718.67 |
Agentic & Tool Use Command A leads
Command A: 35.9 (#40), GPT-5 Nano: 25.8 (#106)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| Berkeley Function Calling Leaderboard | 57.1% | 51.5% |
| Terminal-Bench | — | 21.8% |
Reasoning Command A leads
Command A: 18.3 (#283), GPT-5 Nano: 16.3 (#306)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| Kagi LLM Benchmark | 28.8% | 62.2% |
| LMArena Hard Prompts | 1326 | 1328 |
| DTBench | 61.3% | 62.7% |
| LMCA | 10.3% | 7.9% |
| ARC-AGI-2 | — | 2.6% |
| ARC-AGI-1 | — | 20.7% |
| Chess Puzzles | — | 27% |
| Mystery Game Puzzles | — | 9% |
| Epoch Capabilities Index | — | 139.38 |
| ForecastBench | — | 59.1 |
Math Command A leads
Command A: 36.2 (#171), GPT-5 Nano: 29.4 (#241)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Math | 1300 | 1317 |
| FrontierMath (Tiers 1-3) | — | 20% |
| FrontierMath Tier 4 | — | 2.4% |
| OTIS Mock AIME 2024-2025 | — | 81.1% |
| ProofBench | — | 12% |
| Omni-MATH | — | 54.6% |
| MATH Level 5 | — | 95.2% |
| FrontierMath (Feb 2025 set) | — | 8.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge Command A leads
Command A: 37.1 (#159), GPT-5 Nano: 35.9 (#178)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| Vectara Hallucination Rate | 9.3% | 10.5% |
| LMArena Expert | 1295 | 1321 |
| GPQA Diamond | — | 69.4% |
| SimpleQA Verified | — | 11.7% |
| MMLU-Pro | — | 77.8% |
| GPQA (HELM) | — | 67.9% |
Multimodal Not comparable
Command A: —, GPT-5 Nano: 31.3 (#108)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Vision | — | 1159 |
| VPCT | — | 37.2% |
Multilingual Too close to call
Command A: 45.3 (#170), GPT-5 Nano: 45.3 (#172)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Non-English | 1313 | 1313 |
| LMArena Chinese | 1327 | 1356 |
| LMArena German | 1341 | 1327 |
| LMArena Japanese | 1285 | 1226 |
| LMArena Korean | 1285 | 1269 |
| LMArena Russian | 1314 | 1296 |
| LMArena Spanish | 1347 | 1360 |
| LMArena French | 1351 | — |
Instruction Following GPT-5 Nano leads
Command A: 69.1 (#177), GPT-5 Nano: 75.0 (#79)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Instruction Following | 1309 | 1306 |
| IFEval | — | 93.2% |
Long Context Command A leads
Command A: 40.6 (#151), GPT-5 Nano: 31.3 (#281)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Longer Query | 1334 | 1312 |
| Fiction.LiveBench | — | 44.4% |
Writing & Preference Command A leads
Command A: 47.6 (#208), GPT-5 Nano: 39.1 (#249)
| Benchmark | Command A | GPT-5 Nano |
|---|---|---|
| LMArena Text | 1331 | 1320 |
| LMArena Creative Writing | 1319 | 1249 |
| EQ-Bench Creative Writing | 1145 | 705 |
| LMArena Multi-Turn | 1339 | 1311 |
| WildBench | — | 80.6% |
Frequently asked questions
Is Command A better than GPT-5 Nano?
Command A is the stronger model overall, scoring 36.5 to 33.5 on the Noometry Index. GPT-5 Nano costs 32× less per token, which makes it the better buy when Command A's lead doesn't matter for your workload.
Which is cheaper, Command A or GPT-5 Nano?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; Command A lists at $2.50 and $10.
Is Command A or GPT-5 Nano better for coding?
GPT-5 Nano scores higher on coding benchmarks: 33.6 versus 27.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 256K.
How many benchmarks do Command A and GPT-5 Nano share?
22 benchmarks have published results for both models. Command A has 24 scored results on Noometry and GPT-5 Nano has 49.