GeekRadar / The comparison desk

COMPARE.

AI models

alibaba

Qwen2.5-Max

AI model Released Jan 25, 2025

Capability & preference

General capability · ECI

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

Qwen2.5-Max 131.2^* Qwen2.5-Max · 95% interval 128.4–133.8

Human preference · Arena

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

Qwen2.5-Max 1366.6 Qwen2.5-Max · 95% interval 1362.5–1370.7 · 32,487 votes · 2026-10-28

Model performance

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

Qwen2.5-Max

Tested configuration Radar task score

Unrated

1/4 core tests · not enough evidence

1 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

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

Qwen2.5-Max 56.1^*% Approx. 95% interval: 50.7–61.6% · 2025-04-01 · 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.

Qwen2.5-Max Not available

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

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

Qwen2.5-Max Not available

Factual answers

SimpleQA Verified · 25%

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

Qwen2.5-Max Not available

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

Qwen2.5-Max

1,381.67 rating

Confidence interval: 1,368.88 – 1,394.46

2073 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,359.00 rating

Confidence interval: 1,350.84 – 1,367.16

5089 votes

Oct 28, 2026

Category: coding
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,338.38 rating

Confidence interval: 1,329.82 – 1,346.94

5012 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,375.62 rating

Confidence interval: 1,370.50 – 1,380.73

18207 votes

Oct 28, 2026

Category: english
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,336.72 rating

Confidence interval: 1,330.78 – 1,342.66

21863 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,336.50 rating

Confidence interval: 1,322.77 – 1,350.23

1684 votes

Oct 28, 2026

Category: expert
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,396.59 rating

Confidence interval: 1,364.39 – 1,428.78

343 votes

Oct 28, 2026

Category: french
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,349.62 rating

Confidence interval: 1,328.58 – 1,370.65

782 votes

Oct 28, 2026

Category: german
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,359.50 rating

Confidence interval: 1,353.25 – 1,365.74

9650 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,361.67 rating

Confidence interval: 1,353.82 – 1,369.53

5710 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,349.24 rating

Confidence interval: 1,339.33 – 1,359.15

3660 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,326.60 rating

Confidence interval: 1,318.89 – 1,334.32

5988 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,374.83 rating

Confidence interval: 1,360.75 – 1,388.91

1857 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,381.28 rating

Confidence interval: 1,373.18 – 1,389.39

5679 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,367.43 rating

Confidence interval: 1,356.91 – 1,377.94

2950 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,382.67 rating

Confidence interval: 1,367.60 – 1,397.74

1594 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,372.00 rating

Confidence interval: 1,365.36 – 1,378.63

8555 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,345.99 rating

Confidence interval: 1,339.21 – 1,352.77

8191 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,334.65 rating

Confidence interval: 1,328.66 – 1,340.65

10959 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,299.34 rating

Confidence interval: 1,277.30 – 1,321.38

734 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,303.57 rating

Confidence interval: 1,275.49 – 1,331.65

468 votes

Oct 28, 2026

Category: korean
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,357.55 rating

Confidence interval: 1,348.95 – 1,366.15

4535 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,369.07 rating

Confidence interval: 1,359.17 – 1,378.97

3307 votes

Oct 28, 2026

Category: math
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,364.44 rating

Confidence interval: 1,355.90 – 1,372.98

4814 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,352.01 rating

Confidence interval: 1,346.63 – 1,357.40

14280 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,366.64 rating

Confidence interval: 1,362.55 – 1,370.73

32487 votes

Oct 28, 2026

Category: overall
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,360.59 rating

Confidence interval: 1,340.20 – 1,380.98

850 votes

Oct 28, 2026

Category: polish
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,352.85 rating

Confidence interval: 1,342.40 – 1,363.31

2985 votes

Oct 28, 2026

Category: russian
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,377.01 rating

Confidence interval: 1,341.49 – 1,412.53

269 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: Qwen2.5-Max
Evaluation type: human preference

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

Qwen2.5-Max

1,381.76 rating

Confidence interval: 1,368.97 – 1,394.55

2073 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,359.17 rating

Confidence interval: 1,351.01 – 1,367.33

5090 votes

Oct 15, 2026

Category: coding
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,338.55 rating

Confidence interval: 1,329.99 – 1,347.11

5014 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,375.57 rating

Confidence interval: 1,370.46 – 1,380.68

18206 votes

Oct 15, 2026

Category: english
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,336.78 rating

Confidence interval: 1,330.84 – 1,342.72

21864 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,336.53 rating

Confidence interval: 1,322.80 – 1,350.26

1684 votes

Oct 15, 2026

Category: expert
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,396.21 rating

Confidence interval: 1,364.02 – 1,428.41

343 votes

Oct 15, 2026

Category: french
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,349.61 rating

Confidence interval: 1,328.58 – 1,370.65

781 votes

Oct 15, 2026

Category: german
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,359.53 rating

Confidence interval: 1,353.29 – 1,365.77

9650 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,361.55 rating

Confidence interval: 1,353.70 – 1,369.41

5709 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,349.28 rating

Confidence interval: 1,339.37 – 1,359.19

3660 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,326.68 rating

Confidence interval: 1,318.97 – 1,334.39

5990 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,374.57 rating

Confidence interval: 1,360.49 – 1,388.64

1857 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,381.29 rating

Confidence interval: 1,373.19 – 1,389.40

5679 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,367.44 rating

Confidence interval: 1,356.92 – 1,377.95

2950 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,382.82 rating

Confidence interval: 1,367.74 – 1,397.89

1593 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,372.14 rating

Confidence interval: 1,365.50 – 1,378.77

8555 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,346.06 rating

Confidence interval: 1,339.28 – 1,352.84

8193 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,334.70 rating

Confidence interval: 1,328.70 – 1,340.70

10960 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,299.66 rating

Confidence interval: 1,277.67 – 1,321.65

735 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,303.60 rating

Confidence interval: 1,275.53 – 1,331.67

468 votes

Oct 15, 2026

Category: korean
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,357.68 rating

Confidence interval: 1,349.08 – 1,366.28

4537 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,369.00 rating

Confidence interval: 1,359.11 – 1,378.90

3308 votes

Oct 15, 2026

Category: math
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,364.47 rating

Confidence interval: 1,355.92 – 1,373.01

4813 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,352.15 rating

Confidence interval: 1,346.77 – 1,357.54

14281 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,366.66 rating

Confidence interval: 1,362.58 – 1,370.75

32487 votes

Oct 15, 2026

Category: overall
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,361.09 rating

Confidence interval: 1,340.70 – 1,381.49

849 votes

Oct 15, 2026

Category: polish
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,352.94 rating

Confidence interval: 1,342.49 – 1,363.40

2986 votes

Oct 15, 2026

Category: russian
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

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

Qwen2.5-Max

1,376.94 rating

Confidence interval: 1,341.42 – 1,412.46

269 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: qwen2.5-max
Evaluation type: human preference

Run it locally

Inference · memory estimate

Qwen2.5-Max

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license: Proprietary

* Provisional data