GeekRadar / The comparison desk
COMPARE.
AI models
tencent

Hy3 preview

AI model Released

Context window
256,000 tokens
Maximum output
64,000 tokens
Open weights
Yes

Capability & preference

Human preference · Arena

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

Hy3 preview 1401.2 Hy3 preview · 95% interval 1393.8–1408.6 · 6,869 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

Hy3 preview

1,435.44 rating

Confidence interval: 1,406.88 – 1,464.00

459 votes

Category
chinese
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

Hy3 preview

1,420.97 rating

Confidence interval: 1,407.57 – 1,434.38

2049 votes

Category
coding
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Hy3 preview

1,338.63 rating

Confidence interval: 1,319.72 – 1,357.54

1047 votes

Category
creative_writing
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

Hy3 preview

1,409.48 rating

Confidence interval: 1,399.17 – 1,419.79

3330 votes

Category
english
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Hy3 preview

1,383.28 rating

Confidence interval: 1,373.46 – 1,393.10

5181 votes

Category
exclude_ties
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

Hy3 preview

1,422.99 rating

Confidence interval: 1,401.13 – 1,444.86

707 votes

Category
expert
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating

Hy3 preview

1,422.50 rating

Confidence interval: 1,382.97 – 1,462.03

233 votes

Category
french
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Hy3 preview

1,412.61 rating

Confidence interval: 1,403.56 – 1,421.67

4579 votes

Category
hard_prompts
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Hy3 preview

1,415.83 rating

Confidence interval: 1,403.32 – 1,428.34

2311 votes

Category
hard_prompts_english
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Hy3 preview

1,398.03 rating

Confidence interval: 1,382.22 – 1,413.84

1429 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Hy3 preview

1,351.07 rating

Confidence interval: 1,334.47 – 1,367.66

1372 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Hy3 preview

1,430.35 rating

Confidence interval: 1,404.67 – 1,456.03

554 votes

Category
industry_legal_and_government
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Hy3 preview

1,415.59 rating

Confidence interval: 1,398.60 – 1,432.58

1243 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Hy3 preview

1,442.43 rating

Confidence interval: 1,415.14 – 1,469.73

452 votes

Category
industry_mathematical
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Hy3 preview

1,405.69 rating

Confidence interval: 1,377.91 – 1,433.47

479 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Hy3 preview

1,423.08 rating

Confidence interval: 1,411.67 – 1,434.49

2794 votes

Category
industry_software_and_it_services
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Hy3 preview

1,369.94 rating

Confidence interval: 1,354.79 – 1,385.10

1601 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Hy3 preview

1,382.81 rating

Confidence interval: 1,370.30 – 1,395.32

2335 votes

Category
instruction_following
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Hy3 preview

1,401.59 rating

Confidence interval: 1,390.28 – 1,412.90

3050 votes

Category
longer_query
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

Hy3 preview

1,424.17 rating

Confidence interval: 1,397.52 – 1,450.82

431 votes

Category
math
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Hy3 preview

1,410.32 rating

Confidence interval: 1,392.76 – 1,427.88

1206 votes

Category
multi_turn
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Hy3 preview

1,389.70 rating

Confidence interval: 1,379.56 – 1,399.83

3539 votes

Category
non_english
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

Hy3 preview

1,401.17 rating

Confidence interval: 1,393.79 – 1,408.56

6869 votes

Category
overall
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

Hy3 preview

1,398.08 rating

Confidence interval: 1,375.42 – 1,420.73

710 votes

Category
russian
Evaluated configuration
Hy3 preview
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

Hy3 preview

1,398.88 rating

Confidence interval: 1,359.00 – 1,438.76

210 votes

Category
spanish
Evaluated configuration
Hy3 preview
Evaluation type
human preference

Run it locally

Inference · memory estimate

Hy3 preview

Not enough verified architecture information for a reliable VRAM estimate.

API cost

USD / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

Hy3 preview

Capabilities

Context window
256,000 tokens
Maximum output
64,000 tokens
Reasoning
Yes
Tool calling
Yes
Open weights
Yes
Model license
tencent-hunyuan-community

Input modalities

  • text

Output modalities

  • text