What is the IQ of opus-5?

opus-5 has an estimated IQ of 135, which ranks #8 out of 173 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 opus-5 best at?

Among the dimensions with published benchmark results, opus-5 ranks highest in Reliability (IQ 132, #6 of 173) and Programmatic Reasoning (IQ 146, #7 of 174). Its lowest-ranked measured dimension is Academic Reasoning (IQ 134, #12 of 199).

What is the EQ of opus-5?

opus-5 has an estimated EQ of 139. EQ measures emotional reasoning — how well a model reads and responds to emotional context — and is scored separately from IQ.

Which benchmarks has opus-5 been tested on?

opus-5 has published results on 32 of the 39 benchmarks behind AI IQ's Composite IQ, including Humanity's Last Exam, AA Omniscience, AA-Briefcase v1.1 Elo, GDPval-AA v2.1 normalized score and CritPt.

How much does opus-5 cost?

opus-5 costs $5.00 per million input tokens and $25.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $44.62 per million input-plus-output tokens.

Is opus-5 open source?

No — opus-5 is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.

Who developed opus-5?

opus-5 was developed by Anthropic, based in the United States.

When was opus-5 released?

opus-5 was publicly released on July 24, 2026.

How large is the context window of opus-5?

opus-5 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.

In independent measurements, opus-5 generates about 58 tokens per second, with a typical response time of about 31 seconds.

AnthropicReleased Jul 24, 2026

Proprietary

opus-5

Strongest in Reliability (#6 of 173) and Programmatic Reasoning (#7 of 174); weakest in Academic Reasoning (#12 of 199). All seven dimensions are measured.

Composite IQ ?

135

#8 of 173 ranked models · ahead of 96%

Anthropic's #4 model, behind opus-5.5 (141)

Effective cost

$44.62/1M

$5 in · $25 out

Speed

58tok/s

31s typical response

Context

1Mtokens

1,000,000 tokens in

Evidence

7of 7 dims

32 of 39 benchmarks

Compare with gpt-6.1-solAPI pricing

IQ profile

Seven dimensions, averaged equally into Composite IQ

Compare shape with

Measured95% 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

125measured

#8 / 197

5/5

-2.5

Mathematical ReasoningMath

144measured

#8 / 173

4/5

-2

Academic ReasoningAcademic

134measured

#12 / 199

5/6

-1

Programmatic ReasoningCoding

146measured

#7 / 174

4/6

-2.5

Computer UseComputer Use

137measured

#8 / 173

4/5

-5

ReliabilityReliability

132measured

#6 / 173

4/6

-0.5

Knowledge WorkKnowledge Work

130measured

#10 / 173

6/6

-1

opus-595% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

32 of 39 Composite IQ benchmarks have a published result for opus-5. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 125 · measuredBest: #6 of 68 on ARC-AGI-25/5

BenchmarkScorePercentileRankLeader · source

ARC-AGI-2

Full source leaderboard reconciliation · Oct 23

90.4%

#6 / 68

Leader gpt-6-astra · 95.0%
Full source leaderboard reconciliation · Oct 23

Mystery Game Puzzles

Full source leaderboard reconciliation vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 23

59.0%

#5 / 55

Leader gpt-6-astra · 84.0%
Full source leaderboard reconciliation vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 23

ARC-AGI-1

Full source leaderboard reconciliation · Oct 23

97.5%

#10 / 68

Leader fable-5 · 98.5%
Full source leaderboard reconciliation · Oct 23

ARC-AGI-3

Full source leaderboard reconciliation · Oct 23

30.2%

#4 / 23

Leader gpt-6-astra · 62.7%
Full source leaderboard reconciliation · Oct 23

Chess Puzzles

Epoch AI vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 14

42.0%

#15 / 76

Leader gpt-6-astra · 72.0%
Epoch AI vEpoch 100-position leaderboard (task implementation versions recorded per run) · Oct 14

Mathematical Reasoning IQ 144 · measuredBest: #6 of 61 on MathArena4/5

BenchmarkScorePercentileRankLeader · source

MathArena

Full source leaderboard reconciliation · Oct 23

70.4%

#6 / 61

Leader gpt-6.1-sol · 84.0%
Full source leaderboard reconciliation · Oct 23

FrontierMath Tier 1-3

Full source leaderboard reconciliation · Oct 23

85.6%

#11 / 83

Leader gpt-6-astra · 93.7%
Full source leaderboard reconciliation · Oct 23

FrontierMath Tier 4

Full source leaderboard reconciliation · Oct 23

73.2%

#10 / 71

Leader gpt-6.1-sol · 100%
Full source leaderboard reconciliation · Oct 23

ProofBench 1.1

Full source leaderboard reconciliation v1.1 · Oct 23

99.0%

#7 / 47

Leader fable-5.1 · 100%
Full source leaderboard reconciliation v1.1 · Oct 23

Not yet reported

AIME

Academic Reasoning IQ 134 · measuredBest: #6 of 165 on Humanity's Last Exam5/6

BenchmarkScorePercentileRankLeader · source

Humanity's Last Exam

Artificial Analysis v4.3.2 · Oct 23

54.9%

#6 / 165

Leader opus-5.5 · 61.4%
Artificial Analysis v4.3.2 · Oct 23

CritPt

Artificial Analysis v4.3.2 · Oct 23

29.1%

#10 / 157

Leader Muse Wonder Full · 32.9%
Artificial Analysis v4.3.2 · Oct 23

MMMU-Pro

Artificial Analysis v4.3.2 · Oct 23

84.7%

#9 / 87

Leader fable-5.1 · 90.6%
Artificial Analysis v4.3.2 · Oct 23

GPQA Diamond

Artificial Analysis v4.3.2 · Oct 23

93.2%

#16 / 147

Leader gpt-6-astra · 96.1%
Artificial Analysis v4.3.2 · Oct 23

SciCode

Artificial Analysis v4.3.2 · Oct 23

56.4%

#26 / 165

Leader opus-5.5 · 66.9%
Artificial Analysis v4.3.2 · Oct 23

Not yet reported

MMLU-Pro

Programmatic Reasoning IQ 146 · measuredBest: #9 of 102 on Terminal-Bench 4.04/6

BenchmarkScorePercentileRankLeader · source

Terminal-Bench 4.0

Artificial Analysis v4.3.2 · Oct 23

49.0%

#9 / 102

Leader sonnet-5.5 · 63.6%
Artificial Analysis v4.3.2 · Oct 23

ProgramBench

Full source leaderboard reconciliation · Oct 23

41.5%

#6 / 61

Leader opus-5.5 · 65.0%
Full source leaderboard reconciliation · Oct 23

SWE-rebench

Full source leaderboard reconciliation · Oct 23

63.4%

#3 / 25

Leader fable-5 · 64.5%
Full source leaderboard reconciliation · Oct 23

IOI V2

Full source leaderboard reconciliation v2 · Oct 23

84.3%

#8 / 42

Leader Gemini 4 Argon · 100%
Full source leaderboard reconciliation v2 · Oct 23

Not yet reported

LiveCodeBenchMirrorCode

Computer Use IQ 137 · measuredBest: #3 of 42 on BrowseComp4/5

BenchmarkScorePercentileRankLeader · source

BrowseComp

Full source leaderboard reconciliation · Oct 23

90.8%

#3 / 42

Leader gpt-6-astra · 91.5%
Full source leaderboard reconciliation · Oct 23

MCP Atlas

Full source leaderboard reconciliation · Oct 23

85.8%

#3 / 32

Leader muse-spark-1.1 · 88.1%
Full source leaderboard reconciliation · Oct 23

Arena.ai Agent Arena

Arena.ai Agent Arena · Oct 23

8.1

#10 / 64

Leader opus-5.5 · 20.9
Arena.ai Agent Arena · Oct 23

Agents' Last Exam

Full source leaderboard reconciliation · Oct 23

32.2%

#5 / 29

Leader opus-5.5 · 38.2%
Full source leaderboard reconciliation · Oct 23

Not yet reported

OSWorld-Verified

Reliability IQ 132 · measuredBest: #7 of 156 on AA Omniscience4/6

BenchmarkScorePercentileRankLeader · source

AA Omniscience

Artificial Analysis v4.3.2 · Oct 23

37.1

#7 / 156

Leader opus-5.5 · 46.4
Artificial Analysis v4.3.2 · Oct 23

BullshitBench v2

Full source leaderboard reconciliation · Oct 23

70.0%

#10 / 83

Leader opus-4.8 · 94.0%
Full source leaderboard reconciliation · Oct 23

SimpleQA Verified

Full source leaderboard reconciliation · Oct 23

59.9%

#19 / 79

Leader Gemini 4 Argon · 76.0%
Full source leaderboard reconciliation · Oct 23

AA Long Context Reasoning v1.1

Artificial Analysis v4.3.2 · Oct 23

79.3%

#56 / 152

Leader kimi-k3 · 88.7%
Artificial Analysis v4.3.2 · Oct 23

Not yet reported

IFBenchFACTS Grounding

Knowledge Work IQ 130 · measuredBest: #6 of 109 on AA-Briefcase v1.1 Elo6/6

BenchmarkScorePercentileRankLeader · source

AA-Briefcase v1.1 Elo

Artificial Analysis v4.3.2 · Oct 23

1,658

#6 / 109

Leader sonnet-5.5 · 1,816
Artificial Analysis v4.3.2 · Oct 23

GDPval-AA v2.1 normalized score

Artificial Analysis v4.3.2 · Oct 20

61.1%

#7 / 124

Leader opus-5.5 · 68.3%
Artificial Analysis v4.3.2 · Oct 20

GDP.pdf

Artificial Analysis v4.3.2 · Oct 23

21.6%

#19 / 104

Leader Muse Wonder Full · 32.6%
Artificial Analysis v4.3.2 · Oct 23

AA-AnalystAgent

Full source leaderboard reconciliation vpass^5-2026-09-29 · Oct 23

53.8%

#7 / 30

Leader gemini-3.7-flash · 60.0%
Full source leaderboard reconciliation vpass^5-2026-09-29 · Oct 23

AutomationBench-AA

Artificial Analysis v4.3.2 · Oct 23

56.6%

#29 / 104

Leader Gemini 4 Argon · 77.5%
Artificial Analysis v4.3.2 · Oct 23

EnterpriseOps-Gym-AA

Artificial Analysis · Sep 29

47.5%

#7 / 19

Leader fable-5 · 51.1%
Artificial Analysis · Sep 29

Release & availability

Facts from the publisher's announcement and documentation.

Released: Jul 24, 2026
Developer: Anthropic
Access: Generally available through the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, Claude Pro, and Claude Max
Weights: ProprietaryHosted access only
Inputs: textimage
Context: 1,000,000 tokens
API price: $5 in · $25 outper 1M tokens · full pricing

What Anthropic says

Adaptive thinking is enabled by default, with effort levels from low through max.
Anthropic reports the largest gains over Opus 4.8 in deep reasoning, agentic coding, long-horizon tasks, and test-time compute scaling.
The model supports a 1M-token context window and up to 128k output tokens.

Limitations

Anthropic reports that Opus 5 remains behind Mythos 5 on cybersecurity tasks.
When thinking is disabled, Anthropic warns that tool calls or internal XML can occasionally appear in visible text.

Publisher statements, not AI IQ measurements.

Sourcesanthropic.com/news/claude-opus-5 platform.claude.com/docs/en/about-claude/models/whats-new-opus-5anthropic.com/claude-opus-5-system-card

About this score

How much of opus-5's IQ is measured, and how the rest is estimated.

32/ 39benchmark results

7/ 7dimensions measured

0estimated

5/5Abstract

4/5Math

5/6Academic

4/6Coding

4/5Computer Use

4/6Reliability

6/6Knowledge 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.

How gaps are handled

Every dimension has at least one published result. Missing benchmarks can also draw on its predecessors, Opus 4.8 and Opus 4.7; a predecessor is an estimation input, not evidence that this model has the predecessor's measured results.

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 opus-5

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

#1 overall

gpt-6.1-sol

IQ141 6 higherEffective cost$6.25 7.1× cheaper

Predecessor

opus-4.8

IQ128 7 lowerEffective cost$42.83 about the same

Anthropic #1

opus-5.5

IQ141 6 higherEffective cost$35.41 1.3× cheaper

Closest IQ

gpt-5.6-sol

IQ134 1 lowerEffective cost$27.22 1.6× cheaper

Most similar profile

gpt-6-sol

IQ137 2 higherEffective cost$9.75 4.6× cheaper

Similar price

gpt-6-astra

IQ140 5 higherEffective cost$33.91 1.3× cheaper

opus-5Comparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about opus-5, 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 opus-5?: opus-5 has an estimated IQ of 135, which ranks #8 out of 173 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 opus-5 best at?: Among the dimensions with published benchmark results, opus-5 ranks highest in Reliability (IQ 132, #6 of 173) and Programmatic Reasoning (IQ 146, #7 of 174). Its lowest-ranked measured dimension is Academic Reasoning (IQ 134, #12 of 199).
What is the EQ of opus-5?: opus-5 has an estimated EQ of 139. EQ measures emotional reasoning — how well a model reads and responds to emotional context — and is scored separately from IQ.
Which benchmarks has opus-5 been tested on?: opus-5 has published results on 32 of the 39 benchmarks behind AI IQ's Composite IQ, including Humanity's Last Exam, AA Omniscience, AA-Briefcase v1.1 Elo, GDPval-AA v2.1 normalized score and CritPt.
How much does opus-5 cost?: opus-5 costs $5.00 per million input tokens and $25.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $44.62 per million input-plus-output tokens.
Is opus-5 open source?: No — opus-5 is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.
Who developed opus-5?: opus-5 was developed by Anthropic, based in the United States.
When was opus-5 released?: opus-5 was publicly released on July 24, 2026.
How large is the context window of opus-5?: opus-5 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.
How fast is opus-5?: In independent measurements, opus-5 generates about 58 tokens per second, with a typical response time of about 31 seconds.