Command A
Command / Cohere
30scores
30benchmarks
$2.5 / $10 per 1M tokenscost in/out
Metadata
Command Closed/API
Aliases: cohere-command-a, cohere-command-a-03-2025, cohere/command-a, cohere/command-a-03-2025, command-a, command-a-03-2025
| Benchmark | Category | Rank | Score | Sampled |
|---|---|---|---|---|
| Berkeley Function-Calling Leaderboard | Agentic | 35 | 46.49% | 2026-05-27 |
| Tau2-Bench Telecom | Agentic | 347 | 15.2% | 2026-05-11 |
| Terminal-Bench Hard | Agentic | 339 | 0.8% | 2026-05-11 |
| LiveCodeBench | Coding | 107 | 35.071% | 2026-05-28 |
| SciCode | Coding | 278 | 28.1% | 2026-05-11 |
| SWE-bench Verified | Coding | 50 | 7.8% | 2026-05-28 |
| Terminal-Bench 2.0 | Coding | 64 | 2.247% | 2026-05-28 |
| GSMA Open Telco Leaderboard | Domain | 40 | 53.50 | 2026-05-06 |
| CorpFin v2 | Finance | 90 | 45.96% | 2026-05-28 |
| Finance Agent v1.1 | Finance | 48 | 4.226% | 2026-05-04 |
| FinanceArena | Finance | 18 | 27.7 | 2026-05-27 |
| TaxEval v2 | Finance | 95 | 61.366% | 2026-05-28 |
| MedQA | Healthcare | 72 | 80.55% | 2026-04-16 |
| HUMAINE | Human Preference | 32 | 3.40 | 2026-05-06 |
| Artificial Analysis Intelligence Index | Intelligence | 358 | 13.48 | 2026-05-11 |
| GPQA Diamond | Intelligence | 96 | 48.485% | 2026-05-28 |
| Humanity's Last Exam | Intelligence | 342 | 4.6% | 2026-05-11 |
| MMLU Pro | Intelligence | 96 | 69.172% | 2026-05-28 |
| MMLU-Pro | Intelligence | 213 | 71.2% | 2026-05-11 |
| CaseLaw v2 | Legal | 13 | 64.523% | 2026-05-04 |
| LegalBench | Legal | 62 | 79.697% | 2026-05-28 |
| AIME | Math | 86 | 13.333% | 2026-04-16 |
| AIME 2025 | Math | 227 | 13% | 2026-05-11 |
| MATH 500 | Math | 44 | 76.2% | 2026-01-09 |
| MGSM | Math | 67 | 85.709% | 2026-01-09 |
| Artificial Analysis Openness Index | Openness | 158 | 33.33 | 2026-05-11 |
| GPQA Diamond | Reasoning | 326 | 52.7% | 2026-05-11 |
| CritPt | Science | 166 | 0% | 2026-05-11 |
| Defects4J | Software Engineering | 24 | 0.261 | 2026-05-27 |
| RepairBench | Software Engineering | 25 | 0.24 | 2026-05-27 |
No matching rows.