What is the IQ of command-a-plus?

command-a-plus has an estimated IQ of 97, which ranks #139 out of 176 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.

What is command-a-plus best at?

Among the dimensions with published benchmark results, command-a-plus ranks highest in Programmatic Reasoning (IQ 104, #108 of 177) and Reliability (IQ 110, #109 of 177). Its lowest-ranked measured dimension is Knowledge Work (IQ 97, #128 of 176).

What is the EQ of command-a-plus?

command-a-plus doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for command-a-plus. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.

Which benchmarks has command-a-plus been tested on?

command-a-plus has published results on 13 of the 39 benchmarks behind AI IQ's Composite IQ, including IFBench, AA Omniscience, Terminal-Bench 4.0, SciCode and CritPt.

How much does command-a-plus cost?

command-a-plus costs $0.00 per million input tokens and $0.00 per million output tokens. AI IQ does not yet have enough task-level token usage data to calculate its effective cost.

Is command-a-plus open source?

Yes — command-a-plus is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.

Who developed command-a-plus?

command-a-plus was developed by Cohere, based in Canada.

When was command-a-plus released?

command-a-plus was publicly released on May 20, 2026.

How large is the context window of command-a-plus?

command-a-plus supports a context window of 192,000 tokens — the amount of text it can consider in a single request.

How fast is command-a-plus?

In independent measurements, command-a-plus generates about 152 tokens per second, with a typical response time of about 17 seconds.

CohereReleased May 20, 2026

Open weights

command-a-plus

Its best results are in Programmatic Reasoning (#108 of 177) and Reliability (#109 of 177); weakest in Knowledge Work (#128 of 176). Three of seven dimensions are estimated, because no results are published yet.

Composite IQ ?

#139 of 176 ranked models · ahead of 21%

World (Other)'s #14 model, behind mistral-large-4 (119)

Effective cost

Not priced

No standard API rate

Speed

152tok/s

17s typical response

Context

192Ktokens

192,000 tokens in

Evidence

4of 7 dims

13 of 39 benchmarks

Compare with gpt-6.1-solAPI pricing

IQ profile

Seven dimensions, averaged equally into Composite IQ

Compare shape with

MeasuredEstimated95% range

#1 in dimensionTop-10 median

IQ dimensions

Ranks are among models scored in each dimension. Ranges are model-based estimates, not calibrated confidence intervals.

DimensionIQPosition · 70–160RankEvidencevs Top 10

Abstract ReasoningAbstract

82estimated

#147 / 201

0/5

-44.5

Mathematical ReasoningMath

87estimated

#159 / 176

0/5

-60

Academic ReasoningAcademic

108measured

#144 / 203

5/6

-27.5

Programmatic ReasoningCoding

104measured

#108 / 177

1/6

-45.5

Computer UseComputer Use

94estimated

#133 / 176

0/5

-49

ReliabilityReliability

110measured

#109 / 177

3/6

-24

Knowledge WorkKnowledge Work

97measured

#128 / 176

4/6

-34.5

command-a-plus95% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

13 of 39 Composite IQ benchmarks have a published result for command-a-plus. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 82 · estimatedNo published results0/5

No published results yet for Abstract Reasoning. The 82 shown is estimated from command-a-plus's other dimensions and will be replaced as results come in. How estimates work

ARC-AGI-3ARC-AGI-2ARC-AGI-1Chess PuzzlesMystery Game Puzzles

Mathematical Reasoning IQ 87 · estimatedNo published results0/5

No published results yet for Mathematical Reasoning. The 87 shown is estimated from command-a-plus's other dimensions and will be replaced as results come in. How estimates work

FrontierMath Tier 4FrontierMath Tier 1-3ProofBench 1.1MathArenaAIME

Academic Reasoning IQ 108 · measuredBest: #128 of 167 on SciCode5/6

BenchmarkScorePercentileRankLeader · source

SciCode

Artificial Analysis v4.3.2 · Oct 27

38.5%

#128 / 167

Leader opus-5.5 · 66.9%
Artificial Analysis v4.3.2 · Oct 27

CritPt

Artificial Analysis v4.3.2 · Oct 27

0.3%

#125 / 159

Leader Muse Wonder Full · 32.9%
Artificial Analysis v4.3.2 · Oct 27

GPQA Diamond

Artificial Analysis v4.3.2 · Oct 27

76.1%

#116 / 147

Leader gpt-6-astra · 96.1%
Artificial Analysis v4.3.2 · Oct 27

Humanity's Last Exam

Artificial Analysis v4.3.2 · Oct 27

12.0%

#137 / 167

Leader opus-5.5 · 61.4%
Artificial Analysis v4.3.2 · Oct 27

MMMU-Pro

Artificial Analysis v4.3.2 · Oct 27

63.2%

#79 / 89

Leader fable-5.1 · 90.6%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

MMLU-Pro

Programmatic Reasoning IQ 104 · measuredBest: #63 of 104 on Terminal-Bench 4.01/6

BenchmarkScorePercentileRankLeader · source

Terminal-Bench 4.0

Artificial Analysis v4.3.2 · Oct 27

0.5%

#63 / 104

Leader Kimi K3.1 Pro · 67.2%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

LiveCodeBenchIOI V2ProgramBenchSWE-rebenchMirrorCode

Computer Use IQ 94 · estimatedNo published results0/5

No published results yet for Computer Use. The 94 shown is estimated from command-a-plus's other dimensions and will be replaced as results come in. How estimates work

BrowseCompOSWorld-VerifiedMCP AtlasArena.ai Agent ArenaAgents' Last Exam

Reliability IQ 110 · measuredBest: #28 of 96 on IFBench3/6

BenchmarkScorePercentileRankLeader · source

IFBench

Artificial Analysis v4.3.2 · Oct 27

73.9%

#28 / 96

Leader minimax-m3 · 82.9%
Artificial Analysis v4.3.2 · Oct 27

AA Omniscience

Artificial Analysis v4.3.2 · Oct 27

-4.0

#77 / 158

Leader opus-5.5 · 46.4
Artificial Analysis v4.3.2 · Oct 27

AA Long Context Reasoning v1.1

Artificial Analysis v4.3.2 · Oct 27

52.7%

#136 / 154

Leader kimi-k3 · 88.7%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

SimpleQA VerifiedBullshitBench v2FACTS Grounding

Knowledge Work IQ 97 · measuredBest: #102 of 124 on GDPval-AA v2.1 normalized score4/6

BenchmarkScorePercentileRankLeader · source

GDPval-AA v2.1 normalized score

Artificial Analysis v4.3.2 · Oct 23

0.7%

#102 / 124

Leader opus-5.5 · 68.3%
Artificial Analysis v4.3.2 · Oct 23

AutomationBench-AA

Artificial Analysis v4.3.2 · Oct 27

1.8%

#90 / 106

Leader Gemini 4 Argon · 77.5%
Artificial Analysis v4.3.2 · Oct 27

AA-Briefcase v1.1 Elo

Artificial Analysis v4.3.2 · Oct 27

356

#95 / 111

Leader sonnet-5.5 · 1,812
Artificial Analysis v4.3.2 · Oct 27

GDP.pdf

Artificial Analysis v4.3.2 · Oct 27

1.0%

#101 / 106

Leader Muse Wonder Full · 32.6%
Artificial Analysis v4.3.2 · Oct 27

Not yet reported

AA-AnalystAgentEnterpriseOps-Gym-AA

Release & availability

Release details on record for this model.

Released: May 20, 2026
Developer: Cohere
Weights: Open weightsDownloadable and self-hostable
Context: 192,000 tokens
API price: $0 in · $0 outper 1M tokens · full pricing

About this score

How much of command-a-plus's IQ is measured, and how the rest is estimated.

13/ 39benchmark results

4/ 7dimensions measured

3estimated

0/5Abstract

0/5Math

5/6Academic

1/6Coding

0/5Computer Use

3/6Reliability

4/6Knowledge Work

Bar height = share of benchmarks with resultsEstimatedNo score

Measured results are shown as published

Every score in the benchmark tables comes from its source unchanged. Composite IQ averages all seven dimensions equally.

Estimates fill the gaps

Abstract Reasoning, Mathematical Reasoning and Computer Use have no published results yet, so their IQ is estimated from command-a-plus's other dimensions.

A missing result is not a zero

Benchmark versions and evaluation harnesses can differ between sources, so small gaps between models may not be meaningful.

Read the scoring methodology →

Compare command-a-plus

Side-by-side pages for the models it is most often weighed against.

#1 overall

gpt-6.1-sol

IQ141 44 higherEffective cost$6.30

World (Other) #1

mistral-large-4

IQ119 22 higherEffective cost$4.69

IQ98 1 higherEffective cost$0.403

#2 overall

opus-5.5

IQ141 44 higherEffective cost$35.73

#3 overall

Gemini 4 Argon

IQ140 43 higher

command-a-plusComparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about command-a-plus, drawn from the same data as the rest of this page.

How AI IQ scores models →Get this data from the API →

What is the IQ of command-a-plus?: command-a-plus has an estimated IQ of 97, which ranks #139 out of 176 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.
What is command-a-plus best at?: Among the dimensions with published benchmark results, command-a-plus ranks highest in Programmatic Reasoning (IQ 104, #108 of 177) and Reliability (IQ 110, #109 of 177). Its lowest-ranked measured dimension is Knowledge Work (IQ 97, #128 of 176).
What is the EQ of command-a-plus?: command-a-plus doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for command-a-plus. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
Which benchmarks has command-a-plus been tested on?: command-a-plus has published results on 13 of the 39 benchmarks behind AI IQ's Composite IQ, including IFBench, AA Omniscience, Terminal-Bench 4.0, SciCode and CritPt.
How much does command-a-plus cost?: command-a-plus costs $0.00 per million input tokens and $0.00 per million output tokens. AI IQ does not yet have enough task-level token usage data to calculate its effective cost.
Is command-a-plus open source?: Yes — command-a-plus is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
Who developed command-a-plus?: command-a-plus was developed by Cohere, based in Canada.
When was command-a-plus released?: command-a-plus was publicly released on May 20, 2026.
How large is the context window of command-a-plus?: command-a-plus supports a context window of 192,000 tokens — the amount of text it can consider in a single request.
How fast is command-a-plus?: In independent measurements, command-a-plus generates about 152 tokens per second, with a typical response time of about 17 seconds.