Skip to content
aiqbench

Cohere · model passport

Cohere: Command A

command-a-03-2025
  • Open weights
  • CC-BY-NC-4.0 licence
  • 111.1B params
  • 256,000 context
  • Released 2025-03-01(from model name)
7/27axes covered
Overall · LMArena
1,354#196 of 396
Best percentile
56.2HHEM Hallucination (Vectara)
Cohort standing
44.3median 45.2

What the data says

Benchmarked on 7 of 27 tracked axes. Every figure below is the model's own published result on that board.

Weakest axes

  • 5.8 pct — Factuality Arena (1,363)

Uneven: 50.4 percentile points between its best and worst axis — 56.2 pct on HHEM Hallucination (Vectara) versus 5.8 pct on Factuality Arena.

  • HHEM Hallucination (Vectara)56.2 pct90.7%Verify →
  • Overall50.8 pct1,354Verify →
  • Instruction following50.3 pct1,342Verify →
  • Hard prompts50.3 pct1,368Verify →
  • Coding47.8 pct1,389Verify →
  • LiveBench32.3 pct44.1%Verify →
  • Factuality Arena5.8 pct1,363Verify →
  • WebDev ArenaNot benchmarked
  • Vision ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • GPQA DiamondNot benchmarked
  • AIME (math)Not benchmarked
  • FrontierMathNot benchmarked
  • SimpleQA VerifiedNot benchmarked
  • ARC-AGI-2Not benchmarked
  • Humanity's Last ExamNot benchmarked
  • SimpleBenchNot benchmarked
  • BFCL v4 (function calling)Not benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • MMLU-Pro (Artificial Analysis)Not benchmarked
  • LiveCodeBench (Artificial Analysis)Not benchmarked
  • IFBench (Artificial Analysis)Not benchmarked
  • SWE-bench VerifiedNot benchmarked
  • Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

How to run it

Run it yourself

RepositoryCohereForAI/c4ai-command-a-03-2025
Parameters111.1B
ArchitectureCohere2ForCausalLM
LicenceCC-BY-NC-4.0
AccessGated — you must accept the repo's terms before downloading

Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "CohereForAI/c4ai-command-a-03-2025"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Ollama

We don't sync Ollama tags, so we won't print a command we can't verify — a wrong one would pull the wrong model. Search the Ollama library →

Unsloth fine-tuning recipes are model-specific and we don't derive them here. Unsloth docs →

Use it hosted

API model IDcohere/command-a
Context length256,000 tokens
Max output8,192 tokens
Input price$2.50 per 1M tokens
Output price$10 per 1M tokens

API parameters this endpoint supports

Frequency penaltyMax tokensPresence penaltyResponse formatSeedStop sequencesStructured outputsTemperatureTop-kTop-p

Example request

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"cohere/command-a","messages":[{"role":"user","content":"Hello"}]}'

View on OpenRouter →

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.

  • command-a-03-2025 stands at 44.3 (raw mean 41.9) — measured on 7 of 27 axes.

    standingtemplateVerify →

  • command-a-03-2025 stands at 44.3 versus a cohort median of 45.2 across 144 models released within 180 days — below the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →