GeekRadar / The comparison desk
COMPARE.
AI models
Not reported

oasst-pythia-12b

Model evaluation

Capability & preference

Human preference · Arena

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

oasst-pythia-12b 916.0 oasst-pythia-12b · 95% interval 905.3–926.7 · 6,310 votes · 2026-10-15

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 · 2026-10-15 · rating

oasst-pythia-12b

802.95 rating

Confidence interval: 770.34 – 835.56

321 votes

Category
chinese
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating

oasst-pythia-12b

900.13 rating

Confidence interval: 875.64 – 924.62

714 votes

Category
coding
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

oasst-pythia-12b

923.15 rating

Confidence interval: 903.23 – 943.07

1094 votes

Category
creative_writing
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating

oasst-pythia-12b

945.35 rating

Confidence interval: 933.67 – 957.03

5212 votes

Category
english
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

oasst-pythia-12b

671.53 rating

Confidence interval: 655.64 – 687.42

4325 votes

Category
exclude_ties
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

oasst-pythia-12b

884.44 rating

Confidence interval: 865.90 – 902.98

1153 votes

Category
hard_prompts
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

oasst-pythia-12b

897.38 rating

Confidence interval: 877.16 – 917.60

1002 votes

Category
hard_prompts_english
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

oasst-pythia-12b

822.39 rating

Confidence interval: 792.30 – 852.49

454 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

oasst-pythia-12b

885.28 rating

Confidence interval: 864.61 – 905.95

1067 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

oasst-pythia-12b

922.19 rating

Confidence interval: 882.00 – 962.38

266 votes

Category
industry_legal_and_government
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

oasst-pythia-12b

896.58 rating

Confidence interval: 875.90 – 917.25

1134 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

oasst-pythia-12b

890.65 rating

Confidence interval: 865.77 – 915.52

629 votes

Category
industry_mathematical
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

oasst-pythia-12b

819.07 rating

Confidence interval: 771.94 – 866.20

201 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

oasst-pythia-12b

881.34 rating

Confidence interval: 862.77 – 899.91

1409 votes

Category
industry_software_and_it_services
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

oasst-pythia-12b

928.21 rating

Confidence interval: 910.71 – 945.72

1408 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

oasst-pythia-12b

887.24 rating

Confidence interval: 871.70 – 902.78

1744 votes

Category
instruction_following
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating

oasst-pythia-12b

891.82 rating

Confidence interval: 869.99 – 913.64

687 votes

Category
math
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

oasst-pythia-12b

869.24 rating

Confidence interval: 844.17 – 894.30

683 votes

Category
multi_turn
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating

oasst-pythia-12b

848.11 rating

Confidence interval: 830.30 – 865.91

1098 votes

Category
non_english
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating

oasst-pythia-12b

916.02 rating

Confidence interval: 905.35 – 926.70

6310 votes

Category
overall
Evaluated configuration
oasst-pythia-12b
Evaluation type
human preference

Run it locally

Inference · memory estimate

oasst-pythia-12b

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Apache 2.0