What is the IQ of fable-5.2?

fable-5.2 has an estimated IQ of 144, which ranks #1 out of 180 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 fable-5.2 best at?

Among the dimensions with published benchmark results, fable-5.2 ranks highest in Reliability (IQ 145, #1 of 181) and Computer Use (IQ 157, #1 of 180). Its lowest-ranked measured dimension is Knowledge Work (IQ 134, #4 of 180).

What is the EQ of fable-5.2?

fable-5.2 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 fable-5.2. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.

Which benchmarks has fable-5.2 been tested on?

fable-5.2 has published results on 14 of the 39 benchmarks behind AI IQ's Composite IQ, including FrontierMath Tier 1-3, SimpleQA Verified, FrontierMath Tier 4, Arena.ai Agent Arena and AA-Briefcase v1.1 Elo.

How much does fable-5.2 cost?

fable-5.2 costs $8.00 per million input tokens and $40.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $73.69 per million input-plus-output tokens.

Is fable-5.2 open source?

No — fable-5.2 is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.

Who developed fable-5.2?

fable-5.2 was developed by Anthropic, based in the United States.

When was fable-5.2 released?

fable-5.2 was publicly released on November 9, 2026.

How large is the context window of fable-5.2?

fable-5.2 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.

How fast is fable-5.2?

AI IQ doesn't have independent speed measurements for fable-5.2 yet. Speed comes from independent tests of a model's hosted API, which aren't available for every model — especially new, restricted-access, or self-hosted ones.

AnthropicReleased Nov 9, 2026

NewProprietary

fable-5.2

Strongest in Reliability (#1 of 181) and Computer Use (#1 of 180); weakest in Knowledge Work (#4 of 180). Academic Reasoning is estimated, because no results are published yet.

Composite IQ ?

144

#1 of 180 ranked models · ahead of 100%

Anthropic's top-ranked model

Effective cost

$73.69/1M

$8 in · $40 out

Speed

Not measured

No speed data yet

Context

1Mtokens

1,000,000 tokens in

Evidence

6of 7 dims

14 of 39 benchmarks

Compare with Gemini 4 CarbonAPI 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

134measured

#2 / 209

3/5

Mathematical ReasoningMath

148measured

#3 / 180

3/5

Academic ReasoningAcademic

135estimated

#7 / 211

0/6

Programmatic ReasoningCoding

154measured

#2 / 181

2/6

Computer UseComputer Use

157measured

#1 / 180

1/5

+12.5

ReliabilityReliability

145measured

#1 / 181

2/6

+10.5

Knowledge WorkKnowledge Work

134measured

#4 / 180

3/6

fable-5.295% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human

Benchmark results

14 of 39 Composite IQ benchmarks have a published result for fable-5.2. Benchmarks without a result stay listed so gaps are visible.

Abstract Reasoning IQ 134 · measuredBest: #2 of 71 on ARC-AGI-13/5

BenchmarkScorePercentileRankLeader · source

ARC-AGI-1

ARC Prize · Nov 10

98.5%

#2 / 71

Leader fable-5 · 98.5%
ARC Prize · Nov 10

ARC-AGI-3

ARC Prize · Nov 10

81.2%

#1 / 26

Leader on this benchmark
ARC Prize · Nov 10

ARC-AGI-2

ARC Prize · Nov 10

93.8%

#4 / 71

Leader gpt-6-astra · 95.0%
ARC Prize · Nov 10

Not yet reported

Chess PuzzlesMystery Game Puzzles

Mathematical Reasoning IQ 148 · measuredBest: #1 of 87 on FrontierMath Tier 1-33/5

BenchmarkScorePercentileRankLeader · source

FrontierMath Tier 1-3

Epoch AI · Nov 10

95.1%

#1 / 87

Leader on this benchmark
Epoch AI · Nov 10

FrontierMath Tier 4

Epoch AI · Nov 10

100%

#1 / 75

Leader on this benchmark
Epoch AI · Nov 10

ProofBench 1.1

Vals.ai · Nov 10

100%

#2 / 49

Leader fable-5.1 · 100%
Vals.ai · Nov 10

Not yet reported

MathArenaAIME

Academic Reasoning IQ 135 · estimatedNo published results0/6

No published results yet for Academic Reasoning. The 135 shown is estimated from fable-5.2's other dimensions and will be replaced as results come in. How estimates work

Humanity's Last ExamGPQA DiamondCritPtSciCodeMMLU-ProMMMU-Pro

Programmatic Reasoning IQ 154 · measuredBest: #2 of 63 on ProgramBench2/6

BenchmarkScorePercentileRankLeader · source

ProgramBench

Full source leaderboard reconciliation · Nov 13

63.5%

#2 / 63

Leader opus-5.5 · 65.0%
Full source leaderboard reconciliation · Nov 13

Terminal-Bench 4.0

Artificial Analysis v5.0.1 · Nov 13

59.6%

#4 / 109

Leader Kimi K3.1 Pro · 67.2%
Artificial Analysis v5.0.1 · Nov 13

Not yet reported

LiveCodeBenchIOI V2SWE-rebenchMirrorCode

Computer Use IQ 157 · measuredBest: #1 of 72 on Arena.ai Agent Arena1/5

BenchmarkScorePercentileRankLeader · source

Arena.ai Agent Arena

Arena.ai Agent Arena · Nov 13

30.4

#1 / 72

Leader on this benchmark
Arena.ai Agent Arena · Nov 13

Not yet reported

BrowseCompOSWorld-VerifiedMCP AtlasAgents' Last Exam

Reliability IQ 145 · measuredBest: #1 of 83 on SimpleQA Verified2/6

BenchmarkScorePercentileRankLeader · source

SimpleQA Verified

Epoch AI · Nov 10

94.0%

#1 / 83

Leader on this benchmark
Epoch AI · Nov 10

BullshitBench v2

Full source leaderboard reconciliation · Nov 13

66.0%

#15 / 88

Leader opus-4.8 · 94.0%
Full source leaderboard reconciliation · Nov 13

Not yet reported

aaOmniscienceIFBenchAA Long Context Reasoning v1.1FACTS Grounding

Knowledge Work IQ 134 · measuredBest: #3 of 116 on AA-Briefcase v1.1 Elo3/6

BenchmarkScorePercentileRankLeader · source

AA-Briefcase v1.1 Elo

Artificial Analysis v5.0.1 · Nov 13

1,785

#3 / 116

Leader sonnet-5.5 · 1,816
Artificial Analysis v5.0.1 · Nov 13

AutomationBench-AA

Artificial Analysis v5.0.1 · Nov 13

76.4%

#3 / 112

Leader Gemini 4 Carbon · 77.8%
Artificial Analysis v5.0.1 · Nov 13

GDP.pdf

Artificial Analysis v5.0.1 · Nov 13

24.2%

#16 / 111

Leader Muse Wonder Full · 32.6%
Artificial Analysis v5.0.1 · Nov 13

Not yet reported

GDPval-AA v2.1AA-AnalystAgentEnterpriseOps-Gym-AA

Release & availability

Facts from the publisher's announcement and documentation.

Released: Nov 9, 2026
Developer: Anthropic
Access: Available on the Claude API, Claude apps, Amazon Web Services, Google Cloud, and Microsoft Azure
Weights: ProprietaryHosted access only
Inputs: textimage
Context: 1,000,000 tokens
API price: $8 in · $40 outper 1M tokens · full pricing

What Anthropic says

Anthropic released Fable 5.2 for demanding reasoning, coding, research, and long-horizon agentic work.
Adaptive thinking is always on; effort controls thinking depth and high is the API default.
Anthropic reports faster output than Fable 5.1 and offers a separately priced optional fast mode.

Limitations

Cybersecurity and biology safeguards can block some tasks; verified organizations may request Mythos 5.2 for those uses.
Anthropic requires 30-day retention for safety monitoring unless it expressly authorizes zero data retention.

Publisher statements, not AI IQ measurements.

Sourcesanthropic.com/claude-fable-and-mythos-5-2 platform.claude.com/docs/en/models/fable-5-2/overviewplatform.claude.com/docs/en/release-notes/overview

About this score

How much of fable-5.2's IQ is measured, and how the rest is estimated.

14/ 39benchmark results

6/ 7dimensions measured

1estimated

3/5Abstract

3/5Math

0/6Academic

2/6Coding

1/5Computer Use

2/6Reliability

3/6Knowledge 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

Academic Reasoning has no published results yet, so its IQ is estimated from fable-5.2's other dimensions. Missing benchmarks can also draw on its predecessor, Fable 5.1; 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 fable-5.2

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

IQ138 6 lowerEffective cost$92.11 1.2× pricier

Most similar profile

gpt-6.1-sol

IQ141 3 lowerEffective cost$6.34 11.6× cheaper

Similar price

gpt-6-astra

IQ139 5 lowerEffective cost$34.62 2.1× cheaper

#4 overall

opus-5.5

IQ140 4 lowerEffective cost$35.97 2.0× cheaper

#5 overall

Gemini 4 Argon

IQ139 5 lower

fable-5.2Comparison model

Compare with any modelBrowse all comparisons →

Frequently asked questions

Quick answers about fable-5.2, 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 fable-5.2?: fable-5.2 has an estimated IQ of 144, which ranks #1 out of 180 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 fable-5.2 best at?: Among the dimensions with published benchmark results, fable-5.2 ranks highest in Reliability (IQ 145, #1 of 181) and Computer Use (IQ 157, #1 of 180). Its lowest-ranked measured dimension is Knowledge Work (IQ 134, #4 of 180).
What is the EQ of fable-5.2?: fable-5.2 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 fable-5.2. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
Which benchmarks has fable-5.2 been tested on?: fable-5.2 has published results on 14 of the 39 benchmarks behind AI IQ's Composite IQ, including FrontierMath Tier 1-3, SimpleQA Verified, FrontierMath Tier 4, Arena.ai Agent Arena and AA-Briefcase v1.1 Elo.
How much does fable-5.2 cost?: fable-5.2 costs $8.00 per million input tokens and $40.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $73.69 per million input-plus-output tokens.
Is fable-5.2 open source?: No — fable-5.2 is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.
Who developed fable-5.2?: fable-5.2 was developed by Anthropic, based in the United States.
When was fable-5.2 released?: fable-5.2 was publicly released on November 9, 2026.
How large is the context window of fable-5.2?: fable-5.2 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.
How fast is fable-5.2?: AI IQ doesn't have independent speed measurements for fable-5.2 yet. Speed comes from independent tests of a model's hosted API, which aren't available for every model — especially new, restricted-access, or self-hosted ones.