GPT-5.2 Pro Benchmark Details

GPT-5.2 Pro currently shows benchmark results led by GPQA Diamond (33 / 461, score 93.20), HLE (55 / 568, score 50), FrontierMath - Tier 4 (9 / 80, score 31.30). This page also compares it with 2 competitor models and 2 predecessor or same-series models, including performance and pricing views when available. 1 source link is attached for reference.

Benchmark Results

GPT-5.2 Pro

Benchmark Results

Abstract Generalization

7 evaluations

Benchmark / mode

Score

Rank/total

ARC-AGI-1

MediumTools

81.17

88 / 176

ARC-AGI-1

Thinking Mode

90.50

60 / 176

ARC-AGI-1

HighTools

85.67

84 / 176

ARC-AGI-1

Extra-High

90.50

60 / 176

ARC-AGI-2

MediumTools

38.47

89 / 164

ARC-AGI-2

Thinking Mode

54.16

76 / 164

ARC-AGI-2

High

54.16

76 / 164

Knowledge Exams

2 evaluations

Benchmark / mode

Score

Rank/total

HLE

Thinking Mode

36.60

165 / 568

HLE

Thinking ModeTools

55 / 568

Scientific Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Thinking Mode

93.20

33 / 461

Commonsense

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleBench

Extra-High

57.40

39 / 93

Mathematics

4 evaluations

Benchmark / mode

Score

Rank/total

FrontierMath v2

Extra-High

12 / 58

FrontierMath Tier 4 v2

Extra-High

14 / 42

FrontierMath - Tier 4

Standard ModeToolsInternet

31.30

9 / 80

FrontierMath - Tier 4

Thinking Mode

31.30

9 / 80

Fact Finding

2 evaluations

Benchmark / mode

Score

Rank/total

BrowseComp

Thinking ModeTools

77.90

23 / 58

BrowseComp

Extra-HighTools

77.90

23 / 58

Capability Indices

1 evaluations

Benchmark / mode

Score

Rank/total

ECI

unknown

155.41

23 / 167

Compare with other models

Competitor Comparison

Benchmark scores for GPT-5.2 Pro compared against top models in its class

GPT-5.2 Pro

Opus 4.5

Gemini 3 Deep Think

Benchmark categories:

The chart shows each model’s highest score per benchmark within the current filter. Out-of-100 benchmarks use raw heights; out-of-range benchmarks are scaled within that benchmark while labels keep the original scores.

9 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.

Benchmark	GPT-5.2 ProCurrent	Opus 4.5
ARC-AGI-1 Score (%) Abstract Generalization	90.50Thinking Level · Extra High	80.00Extended Thinking
ARC-AGI-2 Score (% solved); cost per task (USD) Abstract Generalization	54.16Thinking Level · High	37.64Extended Thinking
HLE Accuracy Knowledge Exams	50.00Thinking Enabled ｜ Tools	43.20Extended Thinking ｜ Tools
GPQA Diamond Accuracy Scientific Reasoning	93.20Thinking Enabled	87.00Extended Thinking
SimpleBench Score (AVG@5) Commonsense	57.40Thinking Level · Extra High	62.00Extended Thinking
FrontierMath - Tier 4 Accuracy Mathematics	31.30Standard Mode ｜ Tools	4.20Standard Mode
FrontierMath Tier 4 v2 Accuracy (verification_code) Mathematics	46.00Thinking Level · Extra High	4.8832K
FrontierMath v2 Accuracy (verification_code) Mathematics	74.00Thinking Level · Extra High	34.3932K
ECI ECI score (capability index, higher is better) Capability Indices	155.41Thinking Level · High	150.10Thinking Level · High

Standard API Pricing: GPT-5.2 Pro vs. Peer Models

Shows standard text input and output pricing side by side for each model. If extended-context pricing exists, the chart keeps the base rate and explains the threshold below.

Source: DataLearnerAI. Standard text prices shown here use the default supplier. · USD / 1M tokens

Model	Supplier	Standard input	Standard output	Base price applies to
GPT-5.2 Pro	OpenAI	$21 / 1M tokens	$168 / 1M tokens	—
Opus 4.5	Anthropic	$5 / 1M tokens	$25 / 1M tokens	—

Version History

How each version of the GPT-5.2 Pro series stacks up on benchmark tests

GPT-5.2 ProGPT-5.1 Pro GPT-5-Pro

Benchmark categories:

9 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.· Click a row to view its trend chart.

Benchmark	GPT-5.2 ProCurrent	GPT-5-Pro
ARC-AGI-1 Score (%) Abstract Generalization	90.50Thinking Level · Extra High	70.17Thinking Enabled
ARC-AGI-2 Score (% solved); cost per task (USD) Abstract Generalization	54.16Thinking Level · High	18.00Thinking Enabled
HLE Accuracy Knowledge Exams	50.00Thinking Enabled ｜ Tools	42.00Thinking Enabled ｜ Tools
GPQA Diamond Accuracy Scientific Reasoning	93.20Thinking Enabled	89.40Thinking Enabled ｜ Tools
SimpleBench Score (AVG@5) Commonsense	57.40Thinking Level · Extra High	61.60Thinking Enabled
FrontierMath - Tier 4 Accuracy Mathematics	31.30Standard Mode ｜ Tools	14.60Standard Mode
FrontierMath Tier 4 v2 Accuracy (verification_code) Mathematics	46.00Thinking Level · Extra High	19.51Thinking Level · High
FrontierMath v2 Accuracy (verification_code) Mathematics	74.00Thinking Level · Extra High	55.79Thinking Level · High
ECI ECI score (capability index, higher is better) Capability Indices	155.41Thinking Level · High	150.26Thinking Level · High

Single-Benchmark Version Trend

Viewing: ARC-AGI-1 · Abstract Generalization

Benchmark

NormalNormal + ToolsThinkingThinking + ToolsDeepDeep + Tools

X-axis shows model and release date, Y-axis shows score; solid lines connect the same mode across versions, while dotted guides align modes within the same generation.

Standard API Pricing Across the GPT-5.2 Pro Series

Shows standard text input and output pricing side by side for each model. If extended-context pricing exists, the chart keeps the base rate and explains the threshold below.

Source: DataLearnerAI. Standard text prices shown here use the default supplier. · USD / 1M tokens

Model	Supplier	Standard input	Standard output	Base price applies to
GPT-5.2 Pro	OpenAI	$21 / 1M tokens	$168 / 1M tokens	—
GPT-5.1 Pro	OpenAI	$15 / 1M tokens	$120 / 1M tokens	—
GPT-5-Pro	OpenAI	$15 / 1M tokens	$120 / 1M tokens	—

Sources

openai.comopenai.com