VerdictPal
Join

Models / Cohere

Command A

Command · Released 2026-03-15

Cohere's enterprise-focused Command tier. Strong at RAG, tool use, and multilingual retrieval workflows. 256K context.

Pick when your stack already uses Cohere embed + rerank and you want one vendor for retrieval pipelines.

#48 of 51 on AA Index · current snapshot
8 AA Index · 14 benchmark rows · 2 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite31
Context256K
Input / 1M$2.5
Output / 1M$10
Knowledge cutoff2025-06-01
Output speed61 t/s
TTFT (default API)0.36s
Statussolid
KnowledgeReasoningMathCodingAgenticLong contextTool use
Knowledge
71
Reasoning
53
Math
82
Coding
29
Agentic
15
Long context
18
Tool use
71

48average across 7 tested families

Benchmark placements

Where Command A places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
LiveCodeBenchCoding#12/ 1329Artificial Analysis2026-07-21
MATHMath#14/ 2082Artificial Analysis2026-07-21
AIMEMath#18/ 1913Artificial Analysis2026-07-21
MMLU-ProKnowledge#20/ 2171Artificial Analysis2026-07-21
τ³-Bench BankingAgentic#46/ 6415Artificial Analysis2026-07-21
AA output speedPerformance#47/ 69Artificial Analysis2026-07-21
AA time to first tokenPerformance#57/ 69Artificial Analysis2026-07-21
SciCodeCoding#61/ 6528Artificial Analysis2026-07-21
AA-LCRLong context#62/ 6418Artificial Analysis2026-07-21
Humanity's Last ExamReasoning#62/ 655Artificial Analysis2026-07-21
Artificial Analysis Intelligence IndexReasoning#65/ 678Artificial Analysis2026-07-21
GPQA DiamondReasoning#66/ 6953Artificial Analysis2026-07-21
Terminal-BenchAgentic#68/ 691Artificial Analysis2026-07-21
Family context

The three highest-scoring carded models in each capability family. Where Command A shows up, it's highlighted.

KnowledgeCommand A · 71
  1. 1
    GPT 5.5
    91
  2. 2
    Claude Opus 4.8
    91
  3. 3
    Claude Sonnet 4.6
    91
ReasoningCommand A · 53
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
MathCommand A · 82
  1. 1
    OpenAI o3
    99
  2. 2
    OpenAI o1
    97
  3. 3
    Gemini 3 Pro
    96
CodingCommand A · 29
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticCommand A · 15
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextCommand A · 18
  1. 1
    Kimi K3
    75
  2. 2
    GPT-5.1
    75
  3. 3
    GPT-5.6 Sol
    74
Tool useCommand A · 71
  1. 1
    Claude Sonnet 4.6
    89
  2. 2
    GPT 5.5
    87
  3. 3
    Claude Opus 4.8
    87
Newest receipts

Benchmark rows added to the public ledger for Command A in the last 120 days. Older rows live in the full table below.

14 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context18%Artificial Analysis2
AA output speedPerformance61 t/sArtificial Analysis2
AA time to first tokenPerformance0.36sArtificial Analysis2
AIMEMath13%Artificial Analysis2
Artificial Analysis Intelligence IndexReasoning7.7%Artificial Analysis2
GPQA DiamondReasoning52.7%Artificial Analysis2
Humanity's Last ExamReasoning4.6%Artificial Analysis2
LiveCodeBenchCoding28.7%Artificial Analysis2
MATHMath81.87%Artificial Analysis2
MMLU-ProKnowledge71.2%Artificial Analysis2
SciCodeCoding28.1%Artificial Analysis2
τ³-Bench BankingAgentic15.2%Artificial Analysis2
Terminal-BenchAgentic0.76%Artificial Analysis2
BFCLTool use70.6%Berkeley Function Calling Leaderboard V499
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Knowledge71
Reasoning53
Math82
Coding29
Agentic15
Long context18
Tool use71
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMiniMaxxAIMetaZhipuXiaomiMeituanCohere

Full benchmark ledger

Every catalog benchmark for Command A. Scores link to the original source; gaps mean no public row exists yet.

14 sourced rows

14 of 32 catalog benchmarks have a sourced row for Command A.

Knowledge

Reasoning

Math

Coding

Agentic

Long context

Tool use

Performance

Sources