GeekRadar / The comparison desk
COMPARE.
AI models
google

Gemini 4 Carbon (High)

Model evaluation

Capability & preference

Human preference · Arena

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

Gemini 4 Carbon (High) 1540.0 Gemini 4 Carbon (High) · 95% interval 1530.7–1549.4 · 4,403 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

Gemini 4 Carbon (High)

1,616.20 rating

Confidence interval: 1,588.16 – 1,644.25

474 votes

Category
chinese
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,569.49 rating

Confidence interval: 1,551.29 – 1,587.69

1106 votes

Category
coding
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,530.87 rating

Confidence interval: 1,510.94 – 1,550.79

990 votes

Category
creative_writing
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,530.29 rating

Confidence interval: 1,514.92 – 1,545.67

1558 votes

Category
english
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,568.81 rating

Confidence interval: 1,556.11 – 1,581.51

3354 votes

Category
exclude_ties
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,596.37 rating

Confidence interval: 1,567.87 – 1,624.87

428 votes

Category
expert
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,552.51 rating

Confidence interval: 1,540.73 – 1,564.28

2745 votes

Category
hard_prompts
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,545.11 rating

Confidence interval: 1,526.02 – 1,564.19

1022 votes

Category
hard_prompts_english
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,522.01 rating

Confidence interval: 1,500.22 – 1,543.80

730 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,526.10 rating

Confidence interval: 1,507.96 – 1,544.24

1175 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,565.35 rating

Confidence interval: 1,528.63 – 1,602.08

284 votes

Category
industry_legal_and_government
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,542.76 rating

Confidence interval: 1,519.31 – 1,566.21

656 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,534.20 rating

Confidence interval: 1,496.82 – 1,571.57

247 votes

Category
industry_mathematical
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,527.32 rating

Confidence interval: 1,490.31 – 1,564.34

274 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,555.23 rating

Confidence interval: 1,540.20 – 1,570.26

1621 votes

Category
industry_software_and_it_services
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,543.18 rating

Confidence interval: 1,525.58 – 1,560.78

1272 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,539.19 rating

Confidence interval: 1,523.37 – 1,555.01

1484 votes

Category
instruction_following
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,553.21 rating

Confidence interval: 1,538.77 – 1,567.65

1882 votes

Category
longer_query
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,536.17 rating

Confidence interval: 1,495.17 – 1,577.16

207 votes

Category
math
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,564.68 rating

Confidence interval: 1,539.50 – 1,589.85

602 votes

Category
multi_turn
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,536.92 rating

Confidence interval: 1,525.37 – 1,548.48

2845 votes

Category
non_english
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,540.04 rating

Confidence interval: 1,530.69 – 1,549.40

4403 votes

Category
overall
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

Gemini 4 Carbon (High)

1,563.81 rating

Confidence interval: 1,539.97 – 1,587.64

647 votes

Category
russian
Evaluated configuration
Gemini 4 Carbon (High)
Evaluation type
human preference

Run it locally

Inference · memory estimate

Gemini 4 Carbon (High)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary