GeekRadar / The comparison desk
COMPARE.
AI models
Not reported

Trinity Large Preview

Model evaluation

Capability & preference

Human preference · Arena

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

Trinity Large Preview 1338.6 Trinity Large Preview · 95% interval 1334.4–1342.8 · 31,233 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

Trinity Large Preview

1,353.23 rating

Confidence interval: 1,339.54 – 1,366.92

2132 votes

Category
chinese
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

Trinity Large Preview

1,380.71 rating

Confidence interval: 1,373.43 – 1,387.98

7976 votes

Category
coding
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Trinity Large Preview

1,319.01 rating

Confidence interval: 1,309.88 – 1,328.13

4766 votes

Category
creative_writing
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,357.92 rating

Confidence interval: 1,352.12 – 1,363.72

13890 votes

Category
english
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Trinity Large Preview

1,297.53 rating

Confidence interval: 1,291.74 – 1,303.31

23711 votes

Category
exclude_ties
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

Trinity Large Preview

1,351.00 rating

Confidence interval: 1,339.38 – 1,362.62

2881 votes

Category
expert
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,374.99 rating

Confidence interval: 1,354.10 – 1,395.88

899 votes

Category
french
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,314.69 rating

Confidence interval: 1,286.12 – 1,343.27

463 votes

Category
german
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Trinity Large Preview

1,349.31 rating

Confidence interval: 1,344.09 – 1,354.53

19471 votes

Category
hard_prompts
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,366.95 rating

Confidence interval: 1,359.99 – 1,373.90

8978 votes

Category
hard_prompts_english
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Trinity Large Preview

1,326.92 rating

Confidence interval: 1,318.68 – 1,335.17

6042 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Trinity Large Preview

1,315.83 rating

Confidence interval: 1,307.61 – 1,324.05

5931 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Trinity Large Preview

1,341.15 rating

Confidence interval: 1,327.97 – 1,354.33

2273 votes

Category
industry_legal_and_government
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Trinity Large Preview

1,352.03 rating

Confidence interval: 1,343.35 – 1,360.72

5176 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,354.01 rating

Confidence interval: 1,339.31 – 1,368.72

1786 votes

Category
industry_mathematical
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Trinity Large Preview

1,343.23 rating

Confidence interval: 1,330.14 – 1,356.31

2345 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Trinity Large Preview

1,370.78 rating

Confidence interval: 1,364.62 – 1,376.94

12005 votes

Category
industry_software_and_it_services
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Trinity Large Preview

1,328.25 rating

Confidence interval: 1,320.78 – 1,335.72

7254 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Trinity Large Preview

1,333.69 rating

Confidence interval: 1,327.02 – 1,340.35

9733 votes

Category
instruction_following
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating

Trinity Large Preview

1,292.62 rating

Confidence interval: 1,252.69 – 1,332.54

220 votes

Category
japanese
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,257.40 rating

Confidence interval: 1,228.70 – 1,286.10

469 votes

Category
korean
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,354.14 rating

Confidence interval: 1,347.72 – 1,360.56

11609 votes

Category
longer_query
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,335.75 rating

Confidence interval: 1,322.03 – 1,349.46

1990 votes

Category
math
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,341.72 rating

Confidence interval: 1,333.12 – 1,350.32

5580 votes

Category
multi_turn
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,318.17 rating

Confidence interval: 1,312.83 – 1,323.50

17343 votes

Category
non_english
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,338.61 rating

Confidence interval: 1,334.37 – 1,342.85

31233 votes

Category
overall
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,288.71 rating

Confidence interval: 1,263.14 – 1,314.28

579 votes

Category
polish
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,323.43 rating

Confidence interval: 1,313.28 – 1,333.57

3840 votes

Category
russian
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

Trinity Large Preview

1,357.47 rating

Confidence interval: 1,338.27 – 1,376.66

1039 votes

Category
spanish
Evaluated configuration
Trinity Large Preview
Evaluation type
human preference

Run it locally

Inference · memory estimate

Trinity Large Preview

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Apache 2.0