GeekRadar / The comparison desk
COMPARE.
AI models
alibaba

Qwen3-235B-A22B-Thinking (2507)

Model evaluation

Capability & preference

Human preference · Arena

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

Qwen3-235B-A22B-Thinking (2507) 1415.3 Qwen3-235B-A22B-Thinking (2507) · 95% interval 1408.8–1421.8 · 8,822 votes · 2026-10-28

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

Qwen3-235B-A22B-Thinking (2507)

1,480.78 rating

Confidence interval: 1,454.79 – 1,506.76

536 votes

Category
chinese
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,424.17 rating

Confidence interval: 1,409.47 – 1,438.87

1591 votes

Category
coding
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,383.52 rating

Confidence interval: 1,365.83 – 1,401.20

1083 votes

Category
creative_writing
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,429.66 rating

Confidence interval: 1,420.24 – 1,439.08

3899 votes

Category
english
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,401.79 rating

Confidence interval: 1,392.46 – 1,411.11

6208 votes

Category
exclude_ties
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,462.80 rating

Confidence interval: 1,434.69 – 1,490.92

423 votes

Category
expert
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,408.12 rating

Confidence interval: 1,368.48 – 1,447.76

211 votes

Category
german
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,415.97 rating

Confidence interval: 1,406.51 – 1,425.43

3806 votes

Category
hard_prompts
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,431.86 rating

Confidence interval: 1,417.80 – 1,445.91

1716 votes

Category
hard_prompts_english
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,412.44 rating

Confidence interval: 1,397.42 – 1,427.47

1527 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,369.78 rating

Confidence interval: 1,355.15 – 1,384.41

1528 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,415.00 rating

Confidence interval: 1,390.46 – 1,439.54

532 votes

Category
industry_legal_and_government
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,433.24 rating

Confidence interval: 1,418.01 – 1,448.47

1493 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,435.04 rating

Confidence interval: 1,409.37 – 1,460.71

476 votes

Category
industry_mathematical
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,456.25 rating

Confidence interval: 1,430.38 – 1,482.12

509 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,430.78 rating

Confidence interval: 1,419.86 – 1,441.70

2910 votes

Category
industry_software_and_it_services
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,387.51 rating

Confidence interval: 1,374.03 – 1,401.00

1860 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,385.87 rating

Confidence interval: 1,373.45 – 1,398.29

2095 votes

Category
instruction_following
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,386.56 rating

Confidence interval: 1,357.16 – 1,415.97

442 votes

Category
japanese
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,374.85 rating

Confidence interval: 1,333.88 – 1,415.82

204 votes

Category
korean
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,399.26 rating

Confidence interval: 1,385.11 – 1,413.41

1640 votes

Category
longer_query
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,412.20 rating

Confidence interval: 1,387.54 – 1,436.85

479 votes

Category
math
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,406.42 rating

Confidence interval: 1,390.74 – 1,422.10

1357 votes

Category
multi_turn
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,397.26 rating

Confidence interval: 1,388.88 – 1,405.63

4915 votes

Category
non_english
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,415.29 rating

Confidence interval: 1,408.75 – 1,421.82

8822 votes

Category
overall
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,389.87 rating

Confidence interval: 1,369.26 – 1,410.49

783 votes

Category
polish
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,399.89 rating

Confidence interval: 1,376.06 – 1,423.72

530 votes

Category
russian
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

Qwen3-235B-A22B-Thinking (2507)

1,388.66 rating

Confidence interval: 1,347.65 – 1,429.66

185 votes

Category
spanish
Evaluated configuration
Qwen3-235B-A22B-Thinking (2507)
Evaluation type
human preference

Run it locally

Inference · memory estimate

Qwen3-235B-A22B-Thinking (2507)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Apache 2.0