GeekRadar / The comparison desk

COMPARE.

AI models

xai

Grok 4.20 Beta 1

Model evaluation

Capability & preference

Human preference · Arena

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

Grok 4.20 Beta 1 1444.5 Grok 4.20 Beta 1 · 95% interval 1440.0–1449.0 · 27,889 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

Grok 4.20 Beta 1

1,475.81 rating

Confidence interval: 1,461.28 – 1,490.34

1744 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,446.60 rating

Confidence interval: 1,439.11 – 1,454.09

7483 votes

Oct 28, 2026

Category: coding
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,438.10 rating

Confidence interval: 1,428.49 – 1,447.71

4171 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,450.87 rating

Confidence interval: 1,444.83 – 1,456.90

12790 votes

Oct 28, 2026

Category: english
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,440.50 rating

Confidence interval: 1,434.65 – 1,446.36

21163 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,427.19 rating

Confidence interval: 1,414.91 – 1,439.46

2472 votes

Oct 28, 2026

Category: expert
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,455.40 rating

Confidence interval: 1,433.98 – 1,476.82

859 votes

Oct 28, 2026

Category: french
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,463.69 rating

Confidence interval: 1,436.39 – 1,490.99

443 votes

Oct 28, 2026

Category: german
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,440.80 rating

Confidence interval: 1,435.37 – 1,446.23

17448 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,443.58 rating

Confidence interval: 1,436.43 – 1,450.74

8401 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,424.62 rating

Confidence interval: 1,416.14 – 1,433.11

5357 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,417.17 rating

Confidence interval: 1,408.61 – 1,425.73

5377 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,432.89 rating

Confidence interval: 1,419.55 – 1,446.24

2107 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,440.84 rating

Confidence interval: 1,431.79 – 1,449.90

4610 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,433.66 rating

Confidence interval: 1,418.28 – 1,449.05

1520 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,451.82 rating

Confidence interval: 1,438.16 – 1,465.49

2015 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,452.18 rating

Confidence interval: 1,445.78 – 1,458.58

10827 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,427.61 rating

Confidence interval: 1,419.75 – 1,435.47

6388 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,415.20 rating

Confidence interval: 1,408.26 – 1,422.14

8839 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,449.06 rating

Confidence interval: 1,409.12 – 1,488.99

226 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,416.02 rating

Confidence interval: 1,388.12 – 1,443.92

469 votes

Oct 28, 2026

Category: korean
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,426.72 rating

Confidence interval: 1,419.98 – 1,433.46

10449 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,432.29 rating

Confidence interval: 1,417.96 – 1,446.63

1683 votes

Oct 28, 2026

Category: math
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,450.91 rating

Confidence interval: 1,441.58 – 1,460.25

4596 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,432.87 rating

Confidence interval: 1,427.27 – 1,438.46

15099 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,444.48 rating

Confidence interval: 1,439.95 – 1,449.01

27889 votes

Oct 28, 2026

Category: overall
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,430.66 rating

Confidence interval: 1,406.82 – 1,454.49

582 votes

Oct 28, 2026

Category: polish
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,449.37 rating

Confidence interval: 1,438.74 – 1,460.00

3302 votes

Oct 28, 2026

Category: russian
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

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

Grok 4.20 Beta 1

1,438.35 rating

Confidence interval: 1,418.95 – 1,457.74

952 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: Grok 4.20 Beta 1
Evaluation type: human preference

Run it locally

Inference · memory estimate

Grok 4.20 Beta 1

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license: Proprietary