GeekRadar / The comparison desk

COMPARE.

AI models

Llama 3 70B Instruct

Model evaluation

Capability & preference

Human preference · Arena

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

Llama 3 70B Instruct 1220.8 Llama 3 70B Instruct · 95% interval 1217.2–1224.4 · 156,876 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

Llama 3 70B Instruct

1,114.17 rating

Confidence interval: 1,105.04 – 1,123.30

15885 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,205.84 rating

Confidence interval: 1,198.91 – 1,212.76

28126 votes

Oct 28, 2026

Category: coding
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,209.44 rating

Confidence interval: 1,201.90 – 1,216.98

22753 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,277.37 rating

Confidence interval: 1,272.73 – 1,282.02

90553 votes

Oct 28, 2026

Category: english
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,115.33 rating

Confidence interval: 1,109.77 – 1,120.88

103844 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,148.85 rating

Confidence interval: 1,138.07 – 1,159.62

7958 votes

Oct 28, 2026

Category: expert
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,232.52 rating

Confidence interval: 1,212.74 – 1,252.30

2499 votes

Oct 28, 2026

Category: french
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,169.05 rating

Confidence interval: 1,155.23 – 1,182.88

4991 votes

Oct 28, 2026

Category: german
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,194.74 rating

Confidence interval: 1,189.18 – 1,200.30

45956 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,230.32 rating

Confidence interval: 1,223.63 – 1,237.02

29608 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,172.26 rating

Confidence interval: 1,164.16 – 1,180.36

16626 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,195.34 rating

Confidence interval: 1,188.18 – 1,202.50

23202 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,227.88 rating

Confidence interval: 1,217.18 – 1,238.58

8328 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,225.75 rating

Confidence interval: 1,218.46 – 1,233.04

25903 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,213.48 rating

Confidence interval: 1,205.69 – 1,221.27

18728 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,213.96 rating

Confidence interval: 1,202.81 – 1,225.11

8252 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,204.22 rating

Confidence interval: 1,198.30 – 1,210.14

45062 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,206.01 rating

Confidence interval: 1,199.70 – 1,212.32

37706 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,194.02 rating

Confidence interval: 1,188.72 – 1,199.32

56558 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,016.66 rating

Confidence interval: 1,000.62 – 1,032.69

3276 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,016.67 rating

Confidence interval: 999.41 – 1,033.93

5410 votes

Oct 28, 2026

Category: korean
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,173.69 rating

Confidence interval: 1,166.11 – 1,181.28

17500 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,217.85 rating

Confidence interval: 1,210.41 – 1,225.30

20941 votes

Oct 28, 2026

Category: math
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,222.93 rating

Confidence interval: 1,215.57 – 1,230.30

21421 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,141.64 rating

Confidence interval: 1,136.54 – 1,146.74

66323 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,220.77 rating

Confidence interval: 1,217.19 – 1,224.35

156876 votes

Oct 28, 2026

Category: overall
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,196.26 rating

Confidence interval: 1,147.48 – 1,245.05

406 votes

Oct 28, 2026

Category: polish
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,158.52 rating

Confidence interval: 1,150.22 – 1,166.83

14995 votes

Oct 28, 2026

Category: russian
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

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

Llama 3 70B Instruct

1,241.08 rating

Confidence interval: 1,221.41 – 1,260.75

2705 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: Llama 3 70B Instruct
Evaluation type: human preference

Run it locally

Inference · memory estimate

Llama 3 70B Instruct

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license: Llama 3 Community