GeekRadar / The comparison desk

COMPARE.

AI models

openai

GPT-4o (2024-08-06)

AI model Released Aug 6, 2024

Context window: 128,000 tokens
Maximum output: 16,384 tokens
Open weights: No

Capability & preference

Human preference · Arena

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

GPT-4o (2024-08-06) 1282.5 GPT-4o (2024-08-06) · 95% interval 1278.3–1286.7 · 45,499 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-4o (2024-08-06)

Tested configuration Radar task score

Unrated

3/4 core tests · not enough evidence

3 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

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

GPT-4o (2024-08-06) 49.2^*% Approx. 95% interval: 44.1–54.3% · 2025-01-27 · 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-4o (2024-08-06) Not available

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

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

GPT-4o (2024-08-06) 0.4^*% Approx. 95% interval: 0.0–1.0% · 2026-08-27

Factual answers

SimpleQA Verified · 25%

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

GPT-4o (2024-08-06) 26.0^*% Approx. 95% interval: 23.3–28.7% · 2026-08-31

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.

Test version

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

GPT-4o (2024-08-06)

1,254.07 rating

Confidence interval: 1,243.38 – 1,264.76

3778 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,283.23 rating

Confidence interval: 1,275.38 – 1,291.09

7318 votes

Oct 28, 2026

Category: coding
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,274.27 rating

Confidence interval: 1,265.82 – 1,282.73

6756 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,297.42 rating

Confidence interval: 1,292.08 – 1,302.75

25286 votes

Oct 28, 2026

Category: english
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,208.90 rating

Confidence interval: 1,202.45 – 1,215.35

28538 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,240.39 rating

Confidence interval: 1,227.74 – 1,253.04

2349 votes

Oct 28, 2026

Category: expert
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,263.15 rating

Confidence interval: 1,234.77 – 1,291.54

457 votes

Oct 28, 2026

Category: french
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,257.30 rating

Confidence interval: 1,239.59 – 1,275.00

1202 votes

Oct 28, 2026

Category: german
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,263.85 rating

Confidence interval: 1,257.46 – 1,270.24

12655 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,276.83 rating

Confidence interval: 1,269.15 – 1,284.51

8010 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,231.68 rating

Confidence interval: 1,222.55 – 1,240.80

5095 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,267.38 rating

Confidence interval: 1,259.17 – 1,275.58

7043 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,275.28 rating

Confidence interval: 1,262.78 – 1,287.77

2674 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,273.92 rating

Confidence interval: 1,265.89 – 1,281.95

7560 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,283.21 rating

Confidence interval: 1,274.68 – 1,291.75

5805 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,251.73 rating

Confidence interval: 1,237.88 – 1,265.58

2174 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,277.49 rating

Confidence interval: 1,270.81 – 1,284.18

11864 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,289.59 rating

Confidence interval: 1,282.75 – 1,296.44

12147 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,266.52 rating

Confidence interval: 1,260.72 – 1,272.31

18305 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,234.27 rating

Confidence interval: 1,215.20 – 1,253.33

1104 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,218.23 rating

Confidence interval: 1,193.44 – 1,243.02

712 votes

Oct 28, 2026

Category: korean
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,283.16 rating

Confidence interval: 1,274.81 – 1,291.52

6160 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,285.32 rating

Confidence interval: 1,277.25 – 1,293.39

6826 votes

Oct 28, 2026

Category: math
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,278.81 rating

Confidence interval: 1,270.92 – 1,286.70

8201 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,261.54 rating

Confidence interval: 1,256.04 – 1,267.05

20213 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,282.52 rating

Confidence interval: 1,278.33 – 1,286.71

45499 votes

Oct 28, 2026

Category: overall
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,271.69 rating

Confidence interval: 1,262.91 – 1,280.48

5517 votes

Oct 28, 2026

Category: russian
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

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

GPT-4o (2024-08-06)

1,268.91 rating

Confidence interval: 1,241.81 – 1,296.01

546 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: GPT-4o (2024-08-06)
Evaluation type: human preference

Arena text preference / chinesehuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,254.12 rating

Confidence interval: 1,243.42 – 1,264.81

3778 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,283.33 rating

Confidence interval: 1,275.48 – 1,291.19

7318 votes

Oct 15, 2026

Category: coding
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,274.43 rating

Confidence interval: 1,265.98 – 1,282.88

6756 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,297.37 rating

Confidence interval: 1,292.03 – 1,302.70

25286 votes

Oct 15, 2026

Category: english
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,208.91 rating

Confidence interval: 1,202.46 – 1,215.36

28538 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,240.43 rating

Confidence interval: 1,227.78 – 1,253.09

2349 votes

Oct 15, 2026

Category: expert
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,262.84 rating

Confidence interval: 1,234.44 – 1,291.23

457 votes

Oct 15, 2026

Category: french
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / germanhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,257.30 rating

Confidence interval: 1,239.64 – 1,274.95

1202 votes

Oct 15, 2026

Category: german
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,263.84 rating

Confidence interval: 1,257.45 – 1,270.23

12655 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,276.72 rating

Confidence interval: 1,269.04 – 1,284.40

8010 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,231.72 rating

Confidence interval: 1,222.59 – 1,240.84

5095 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,267.44 rating

Confidence interval: 1,259.23 – 1,275.64

7043 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,275.02 rating

Confidence interval: 1,262.53 – 1,287.52

2674 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,273.90 rating

Confidence interval: 1,265.87 – 1,281.93

7560 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,283.21 rating

Confidence interval: 1,274.68 – 1,291.74

5805 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,251.79 rating

Confidence interval: 1,237.94 – 1,265.64

2174 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,277.57 rating

Confidence interval: 1,270.89 – 1,284.26

11864 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,289.66 rating

Confidence interval: 1,282.82 – 1,296.50

12147 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,266.53 rating

Confidence interval: 1,260.73 – 1,272.32

18305 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,234.10 rating

Confidence interval: 1,215.11 – 1,253.09

1104 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,218.27 rating

Confidence interval: 1,193.50 – 1,243.04

712 votes

Oct 15, 2026

Category: korean
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,283.20 rating

Confidence interval: 1,274.84 – 1,291.56

6160 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,285.28 rating

Confidence interval: 1,277.22 – 1,293.35

6826 votes

Oct 15, 2026

Category: math
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,278.81 rating

Confidence interval: 1,270.91 – 1,286.70

8201 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,261.64 rating

Confidence interval: 1,256.14 – 1,267.14

20213 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,282.50 rating

Confidence interval: 1,278.31 – 1,286.69

45499 votes

Oct 15, 2026

Category: overall
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,271.85 rating

Confidence interval: 1,263.06 – 1,280.63

5517 votes

Oct 15, 2026

Category: russian
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-15 · rating

GPT-4o (2024-08-06)

1,268.83 rating

Confidence interval: 1,241.71 – 1,295.95

546 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: gpt-4o-2024-08-06
Evaluation type: human preference

Run it locally

Inference · memory estimate

GPT-4o (2024-08-06)

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

API cost

USD / month

Input tokens / monthOutput tokens / month

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

GPT-4o (2024-08-06)

Provider offer

Capabilities

Context window: 128,000 tokens
Maximum output: 16,384 tokens
Reasoning: No
Tool calling: Yes
Open weights: No
Knowledge cutoff: 2023-09
Model license: Proprietary

Input modalities

text
image

Output modalities

text

* Provisional data