geekradar
Int
Explore GeekRadar

More to explore

Good gear All topics About us Say hello
GeekRadar / The comparison desk
COMPARE.
AI models
openai

GPT-5.2

AI model Released

API mode
Chat*
Context window
400,000 tokens
Maximum output
128,000 tokens
Open weights
No

Capability & preference

General capability · ECI

A cross-benchmark capability estimate. Index points, not a percentage; intervals show uncertainty.

GPT-5.2 153.8* GPT-5.2 · 95% interval 152.2–155.9

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

GPT-5.2 1412.3 GPT-5.2 · 95% interval 1409.1–1415.4 · 83,793 votes · 2026-10-28

Model performance

Radar task score: four tests, equal weight. An experimental summary of these tasks, not a universal intelligence rating.

GPT-5.2

Radar task score

Unrated

3/4 core tests · not enough evidence

4 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

Expert-written science questions; strict multiple-choice scoring. Chance accuracy is 25%.

GPT-5.2 87.9*% Approx. 95% interval: 84.1–91.6% · Older result

Novel problem solving

Mystery Game Puzzles · 25%

Best-next-move puzzles in an undisclosed game variant. A narrow reasoning test, not general intelligence.

GPT-5.2 22.0*% Approx. 95% interval: 13.8–30.2%

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

Private v2 mathematics problems with Python tools. Earlier versions and Tier 4 are excluded.

GPT-5.2 Not available

Factual answers

SimpleQA Verified · 25%

Correct factual answers under a guessing prompt, graded by a model. Not a hallucination or abstention score.

GPT-5.2 32.7*% Approx. 95% interval: 29.8–35.6%

Competition mathematics

OTIS Mock AIME 2024-2025 · Separate result

Competition problems with model-extracted answers checked against an exact key. Earlier grading procedures are excluded.

GPT-5.2 93.9*% Approx. 95% interval: 87.9–99.9% · Older result

Bars show reported task scores. Whiskers are approximate 95% intervals where standard errors are available. Different reasoning budgets are not equal-cost tests. Missing results are never zero.

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Arena text preference / chinesehuman preference · 2026-10-28 · rating

GPT-5.2

1,446.68 rating

Confidence interval: 1,436.51 – 1,456.86

3577 votes

Category
chinese
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,445.25 rating

Confidence interval: 1,436.67 – 1,453.84

5338 votes

Category
chinese
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

GPT-5.2

1,441.62 rating

Confidence interval: 1,435.61 – 1,447.63

12225 votes

Category
coding
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,431.96 rating

Confidence interval: 1,426.84 – 1,437.08

22111 votes

Category
coding
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

GPT-5.2

1,375.95 rating

Confidence interval: 1,368.65 – 1,383.24

7659 votes

Category
creative_writing
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,375.52 rating

Confidence interval: 1,369.49 – 1,381.55

13902 votes

Category
creative_writing
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

GPT-5.2

1,422.25 rating

Confidence interval: 1,417.38 – 1,427.11

21216 votes

Category
english
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,412.66 rating

Confidence interval: 1,408.38 – 1,416.94

37577 votes

Category
english
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

GPT-5.2

1,403.06 rating

Confidence interval: 1,398.36 – 1,407.76

36886 votes

Category
exclude_ties
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,398.33 rating

Confidence interval: 1,394.12 – 1,402.54

62923 votes

Category
exclude_ties
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

GPT-5.2

1,444.69 rating

Confidence interval: 1,435.07 – 1,454.31

3998 votes

Category
expert
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,428.41 rating

Confidence interval: 1,421.08 – 1,435.74

8008 votes

Category
expert
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating

GPT-5.2

1,454.51 rating

Confidence interval: 1,437.56 – 1,471.47

1434 votes

Category
french
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,436.24 rating

Confidence interval: 1,423.12 – 1,449.35

2721 votes

Category
french
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating

GPT-5.2

1,444.39 rating

Confidence interval: 1,423.62 – 1,465.16

831 votes

Category
german
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,429.70 rating

Confidence interval: 1,412.32 – 1,447.07

1276 votes

Category
german
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

GPT-5.2

1,428.13 rating

Confidence interval: 1,423.75 – 1,432.51

29862 votes

Category
hard_prompts
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,426.87 rating

Confidence interval: 1,423.00 – 1,430.74

53209 votes

Category
hard_prompts
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

GPT-5.2

1,430.54 rating

Confidence interval: 1,424.67 – 1,436.41

13037 votes

Category
hard_prompts_english
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,423.05 rating

Confidence interval: 1,418.04 – 1,428.06

24666 votes

Category
hard_prompts_english
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

GPT-5.2

1,413.88 rating

Confidence interval: 1,407.33 – 1,420.42

9755 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,409.58 rating

Confidence interval: 1,404.01 – 1,415.14

16835 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

GPT-5.2

1,378.44 rating

Confidence interval: 1,372.93 – 1,383.95

17729 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,372.27 rating

Confidence interval: 1,365.61 – 1,378.94

9343 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

GPT-5.2

1,427.09 rating

Confidence interval: 1,417.17 – 1,437.01

3773 votes

Category
industry_legal_and_government
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,420.16 rating

Confidence interval: 1,412.20 – 1,428.12

6680 votes

Category
industry_legal_and_government
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

GPT-5.2

1,425.29 rating

Confidence interval: 1,418.31 – 1,432.26

8278 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,421.09 rating

Confidence interval: 1,415.19 – 1,426.99

13804 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

GPT-5.2

1,442.74 rating

Confidence interval: 1,430.80 – 1,454.67

2619 votes

Category
industry_mathematical
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,418.26 rating

Confidence interval: 1,408.88 – 1,427.63

4556 votes

Category
industry_mathematical
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

GPT-5.2

1,424.34 rating

Confidence interval: 1,416.10 – 1,432.58

6180 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,420.92 rating

Confidence interval: 1,410.48 – 1,431.36

3503 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

GPT-5.2

1,434.24 rating

Confidence interval: 1,429.12 – 1,439.37

18399 votes

Category
industry_software_and_it_services
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,432.59 rating

Confidence interval: 1,428.10 – 1,437.09

32389 votes

Category
industry_software_and_it_services
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

GPT-5.2

1,391.80 rating

Confidence interval: 1,385.66 – 1,397.94

11514 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,391.64 rating

Confidence interval: 1,386.43 – 1,396.85

20417 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

GPT-5.2

1,409.31 rating

Confidence interval: 1,403.78 – 1,414.85

15293 votes

Category
instruction_following
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,402.93 rating

Confidence interval: 1,398.21 – 1,407.66

27766 votes

Category
instruction_following
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating

GPT-5.2

1,419.74 rating

Confidence interval: 1,390.28 – 1,449.21

446 votes

Category
japanese
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,395.13 rating

Confidence interval: 1,373.15 – 1,417.10

805 votes

Category
japanese
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating

GPT-5.2

1,374.66 rating

Confidence interval: 1,357.43 – 1,391.90

1302 votes

Category
korean
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,369.63 rating

Confidence interval: 1,347.89 – 1,391.38

786 votes

Category
korean
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

GPT-5.2

1,412.70 rating

Confidence interval: 1,408.03 – 1,417.37

34042 votes

Category
longer_query
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,408.46 rating

Confidence interval: 1,403.03 – 1,413.89

17219 votes

Category
longer_query
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

GPT-5.2

1,439.92 rating

Confidence interval: 1,429.20 – 1,450.64

3122 votes

Category
math
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,417.65 rating

Confidence interval: 1,408.58 – 1,426.71

4645 votes

Category
math
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

GPT-5.2

1,419.17 rating

Confidence interval: 1,412.35 – 1,425.99

9240 votes

Category
multi_turn
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,418.42 rating

Confidence interval: 1,412.75 – 1,424.09

15938 votes

Category
multi_turn
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

GPT-5.2

1,405.32 rating

Confidence interval: 1,401.43 – 1,409.21

46212 votes

Category
non_english
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,404.84 rating

Confidence interval: 1,400.47 – 1,409.21

28412 votes

Category
non_english
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

GPT-5.2

1,416.25 rating

Confidence interval: 1,412.74 – 1,419.77

49628 votes

Category
overall
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,412.28 rating

Confidence interval: 1,409.14 – 1,415.43

83793 votes

Category
overall
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating

GPT-5.2

1,425.91 rating

Confidence interval: 1,410.99 – 1,440.83

1631 votes

Category
polish
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,423.00 rating

Confidence interval: 1,405.93 – 1,440.06

1180 votes

Category
polish
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

GPT-5.2

1,415.30 rating

Confidence interval: 1,408.69 – 1,421.91

9718 votes

Category
russian
Evaluated configuration
GPT-5.2
Evaluation type
human preference

1,411.51 rating

Confidence interval: 1,403.76 – 1,419.27

6312 votes

Category
russian
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

GPT-5.2

1,413.42 rating

Confidence interval: 1,397.86 – 1,428.99

1537 votes

Category
spanish
Evaluated configuration
GPT-5.2 (High)
Evaluation type
human preference

1,399.26 rating

Confidence interval: 1,386.49 – 1,412.04

2564 votes

Category
spanish
Evaluated configuration
GPT-5.2
Evaluation type
human preference
Arena text preference / chinesehuman preference · 2026-10-15 · rating

GPT-5.2

1,446.64 rating

Confidence interval: 1,436.47 – 1,456.82

3575 votes

Category
chinese
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,445.17 rating

Confidence interval: 1,436.59 – 1,453.74

5354 votes

Category
chinese
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating

GPT-5.2

1,441.59 rating

Confidence interval: 1,435.58 – 1,447.60

12245 votes

Category
coding
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,431.82 rating

Confidence interval: 1,426.70 – 1,436.93

22123 votes

Category
coding
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

GPT-5.2

1,375.77 rating

Confidence interval: 1,368.48 – 1,383.07

7660 votes

Category
creative_writing
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,375.47 rating

Confidence interval: 1,369.43 – 1,381.50

13925 votes

Category
creative_writing
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating

GPT-5.2

1,422.31 rating

Confidence interval: 1,417.45 – 1,427.18

21224 votes

Category
english
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,412.65 rating

Confidence interval: 1,408.36 – 1,416.93

37602 votes

Category
english
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

GPT-5.2

1,403.05 rating

Confidence interval: 1,398.35 – 1,407.75

36902 votes

Category
exclude_ties
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,398.19 rating

Confidence interval: 1,393.98 – 1,402.41

62982 votes

Category
exclude_ties
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-15 · rating

GPT-5.2

1,445.22 rating

Confidence interval: 1,435.60 – 1,454.84

4001 votes

Category
expert
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,428.13 rating

Confidence interval: 1,420.79 – 1,435.47

8008 votes

Category
expert
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating

GPT-5.2

1,454.33 rating

Confidence interval: 1,437.38 – 1,471.28

1440 votes

Category
french
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,435.40 rating

Confidence interval: 1,422.25 – 1,448.54

2734 votes

Category
french
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating

GPT-5.2

1,443.97 rating

Confidence interval: 1,423.19 – 1,464.76

834 votes

Category
german
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,429.50 rating

Confidence interval: 1,412.07 – 1,446.94

1277 votes

Category
german
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

GPT-5.2

1,427.98 rating

Confidence interval: 1,423.60 – 1,432.36

29889 votes

Category
hard_prompts
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,426.76 rating

Confidence interval: 1,422.89 – 1,430.63

53250 votes

Category
hard_prompts
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

GPT-5.2

1,430.54 rating

Confidence interval: 1,424.67 – 1,436.41

13040 votes

Category
hard_prompts_english
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,423.00 rating

Confidence interval: 1,417.99 – 1,428.01

24686 votes

Category
hard_prompts_english
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

GPT-5.2

1,414.04 rating

Confidence interval: 1,407.49 – 1,420.59

9758 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,409.60 rating

Confidence interval: 1,404.04 – 1,415.17

16843 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

GPT-5.2

1,378.37 rating

Confidence interval: 1,372.86 – 1,383.88

17744 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,372.07 rating

Confidence interval: 1,365.40 – 1,378.74

9347 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

GPT-5.2

1,427.26 rating

Confidence interval: 1,417.33 – 1,437.19

3769 votes

Category
industry_legal_and_government
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,420.43 rating

Confidence interval: 1,412.47 – 1,428.40

6695 votes

Category
industry_legal_and_government
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

GPT-5.2

1,425.05 rating

Confidence interval: 1,418.07 – 1,432.02

8275 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,421.02 rating

Confidence interval: 1,415.12 – 1,426.93

13798 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

GPT-5.2

1,442.38 rating

Confidence interval: 1,430.45 – 1,454.31

2622 votes

Category
industry_mathematical
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,418.20 rating

Confidence interval: 1,408.81 – 1,427.59

4555 votes

Category
industry_mathematical
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

GPT-5.2

1,424.64 rating

Confidence interval: 1,416.39 – 1,432.89

6179 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,420.80 rating

Confidence interval: 1,410.36 – 1,431.24

3509 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

GPT-5.2

1,434.23 rating

Confidence interval: 1,429.10 – 1,439.35

18421 votes

Category
industry_software_and_it_services
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,432.38 rating

Confidence interval: 1,427.89 – 1,436.87

32419 votes

Category
industry_software_and_it_services
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

GPT-5.2

1,391.76 rating

Confidence interval: 1,385.62 – 1,397.91

11510 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,391.31 rating

Confidence interval: 1,386.10 – 1,396.52

20443 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

GPT-5.2

1,409.23 rating

Confidence interval: 1,403.70 – 1,414.77

15300 votes

Category
instruction_following
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,402.85 rating

Confidence interval: 1,398.12 – 1,407.57

27797 votes

Category
instruction_following
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating

GPT-5.2

1,419.63 rating

Confidence interval: 1,390.12 – 1,449.14

446 votes

Category
japanese
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,395.23 rating

Confidence interval: 1,373.15 – 1,417.32

804 votes

Category
japanese
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating

GPT-5.2

1,374.10 rating

Confidence interval: 1,356.81 – 1,391.38

1302 votes

Category
korean
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,370.16 rating

Confidence interval: 1,348.38 – 1,391.94

788 votes

Category
korean
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

GPT-5.2

1,412.66 rating

Confidence interval: 1,407.98 – 1,417.33

34073 votes

Category
longer_query
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,408.60 rating

Confidence interval: 1,403.17 – 1,414.03

17230 votes

Category
longer_query
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating

GPT-5.2

1,439.81 rating

Confidence interval: 1,429.09 – 1,450.52

3127 votes

Category
math
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,417.75 rating

Confidence interval: 1,408.68 – 1,426.83

4643 votes

Category
math
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

GPT-5.2

1,419.31 rating

Confidence interval: 1,412.49 – 1,426.13

9245 votes

Category
multi_turn
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,418.49 rating

Confidence interval: 1,412.82 – 1,424.16

15945 votes

Category
multi_turn
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating

GPT-5.2

1,405.17 rating

Confidence interval: 1,401.28 – 1,409.07

46249 votes

Category
non_english
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,404.80 rating

Confidence interval: 1,400.43 – 1,409.17

28431 votes

Category
non_english
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating

GPT-5.2

1,416.24 rating

Confidence interval: 1,412.72 – 1,419.76

49655 votes

Category
overall
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,412.14 rating

Confidence interval: 1,408.99 – 1,415.29

83855 votes

Category
overall
Evaluated configuration
gpt-5.2
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating

GPT-5.2

1,425.83 rating

Confidence interval: 1,410.90 – 1,440.77

1631 votes

Category
polish
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,423.25 rating

Confidence interval: 1,406.20 – 1,440.31

1182 votes

Category
polish
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating

GPT-5.2

1,415.08 rating

Confidence interval: 1,408.46 – 1,421.69

9726 votes

Category
russian
Evaluated configuration
gpt-5.2
Evaluation type
human preference

1,411.96 rating

Confidence interval: 1,404.20 – 1,419.72

6310 votes

Category
russian
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating

GPT-5.2

1,413.12 rating

Confidence interval: 1,397.54 – 1,428.70

1536 votes

Category
spanish
Evaluated configuration
gpt-5.2-high
Evaluation type
human preference

1,399.60 rating

Confidence interval: 1,386.79 – 1,412.41

2559 votes

Category
spanish
Evaluated configuration
gpt-5.2
Evaluation type
human preference

Run it locally

Inference · memory estimate

GPT-5.2

Closed weights. No supported local deployment or GPU requirement is published.

API cost

USD / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

GPT-5.2

Capabilities

Context window
400,000 tokens
Maximum input
272,000 tokens*
Maximum output
128,000 tokens
Reasoning
Yes
Tool calling
Yes
Structured output
Yes*
Open weights
No
Knowledge cutoff
2025-08-31
API mode
Chat*
Image input
Yes*
PDF input
Yes*
Parallel tool calls
Yes*
Prompt caching
Yes*
System messages
Yes*
Tool choice
Yes*
Web search
Yes*
Model license
Proprietary

Input modalities

  • text
  • image

Output modalities

  • text
  • image*

System configuration

Detailed specifications

Specifications · 37 fields
cache read input token cost
1.75e-7*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
cache read input token cost batches
8.75e-8*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
cache read input token cost flex
8.75e-8*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
cache read input token cost priority
3.5e-7*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
default reasoning effort
none*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token
1.75e-6*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token batches
8.75e-7*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token flex
8.75e-7*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token priority
3.5e-6*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
litellm provider
openai*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max input tokens
272,000*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max output tokens
128,000*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max tokens
128,000*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
mode
chat*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token
1.4e-5*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token batches
7.0e-6*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token flex
7.0e-6*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token priority
2.8e-5*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
search context cost per query
search_context_size_high
0.01*
search_context_size_low
0.01*
search_context_size_medium
0.01*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supported endpoints
  • /v1/chat/completions*
  • /v1/batch*
  • /v1/responses*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supported modalities
  • text*
  • image*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supported output modalities
  • text*
  • image*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports function calling
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports minimal reasoning effort
No*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports native streaming
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports none reasoning effort
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports parallel function calling
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports pdf input
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports prompt caching
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports reasoning
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports response schema
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports system messages
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports tool choice
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports tool search
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports vision
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports web search
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports xhigh reasoning effort
Yes*
Scope and conditions
native_context
provider
openai*
mode
chat*
full_native_model_id
gpt-5.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*

* Provisional data