mercury
Not enough verified architecture information for a reliable VRAM estimate.
GeekRadar / The comparison desk
COMPARE.
AI modelsmercury
Model evaluation
Capability & preference
Human preference · Arena
Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.
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 / codinghuman preference · 2026-10-15 · rating
mercury
1,322.30 rating
Confidence interval: 1,294.39 – 1,350.21
413 votes
- Category
- coding
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating
mercury
1,189.07 rating
Confidence interval: 1,150.87 – 1,227.28
295 votes
- Category
- creative_writing
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating
mercury
1,300.46 rating
Confidence interval: 1,279.89 – 1,321.03
855 votes
- Category
- english
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating
mercury
1,212.08 rating
Confidence interval: 1,191.59 – 1,232.56
1430 votes
- Category
- exclude_ties
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating
mercury
1,284.79 rating
Confidence interval: 1,266.43 – 1,303.15
1053 votes
- Category
- hard_prompts
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating
mercury
1,307.94 rating
Confidence interval: 1,280.11 – 1,335.78
455 votes
- Category
- hard_prompts_english
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating
mercury
1,290.37 rating
Confidence interval: 1,258.70 – 1,322.03
359 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating
mercury
1,214.19 rating
Confidence interval: 1,182.26 – 1,246.12
376 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating
mercury
1,299.86 rating
Confidence interval: 1,265.84 – 1,333.88
297 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating
mercury
1,319.43 rating
Confidence interval: 1,297.12 – 1,341.75
686 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating
mercury
1,229.03 rating
Confidence interval: 1,200.48 – 1,257.58
466 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating
mercury
1,239.54 rating
Confidence interval: 1,214.56 – 1,264.52
590 votes
- Category
- instruction_following
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating
mercury
1,264.13 rating
Confidence interval: 1,238.44 – 1,289.81
522 votes
- Category
- longer_query
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating
mercury
1,282.29 rating
Confidence interval: 1,250.07 – 1,314.51
333 votes
- Category
- multi_turn
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating
mercury
1,259.20 rating
Confidence interval: 1,241.37 – 1,277.04
1110 votes
- Category
- non_english
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating
mercury
1,281.81 rating
Confidence interval: 1,268.11 – 1,295.51
1965 votes
- Category
- overall
- Evaluated configuration
- mercury
- Evaluation type
- human preference
Run it locally
Capabilities
- Model license
- Proprietary
0 selected Compare 0