GeekRadar / The comparison desk

COMPARE.

AI models

Llama-2-70b-chat

AI model

Open weights: Yes

Capability & preference

Human preference · Arena

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

Llama-2-70b-chat 1115.2 Llama 2 70B Chat · 95% interval 1109.8–1120.6 · 38,492 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.

Test version

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

Llama-2-70b-chat

994.58 rating

Confidence interval: 981.21 – 1,007.95

3619 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,078.49 rating

Confidence interval: 1,068.69 – 1,088.30

5717 votes

Oct 28, 2026

Category: coding
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,074.58 rating

Confidence interval: 1,064.54 – 1,084.63

6376 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,155.98 rating

Confidence interval: 1,149.55 – 1,162.40

26045 votes

Oct 28, 2026

Category: english
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

955.08 rating

Confidence interval: 947.03 – 963.13

25547 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,039.06 rating

Confidence interval: 1,021.88 – 1,056.23

1356 votes

Oct 28, 2026

Category: expert
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,090.38 rating

Confidence interval: 1,062.38 – 1,118.38

679 votes

Oct 28, 2026

Category: french
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,041.27 rating

Confidence interval: 1,018.84 – 1,063.70

878 votes

Oct 28, 2026

Category: german
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,072.90 rating

Confidence interval: 1,064.78 – 1,081.01

9650 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,091.26 rating

Confidence interval: 1,082.09 – 1,100.43

6989 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,082.76 rating

Confidence interval: 1,070.81 – 1,094.71

3543 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,064.37 rating

Confidence interval: 1,054.47 – 1,074.27

6070 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,114.04 rating

Confidence interval: 1,098.58 – 1,129.50

2073 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,119.75 rating

Confidence interval: 1,109.79 – 1,129.71

6638 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,089.43 rating

Confidence interval: 1,078.42 – 1,100.44

4052 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,093.11 rating

Confidence interval: 1,077.70 – 1,108.52

2169 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,098.86 rating

Confidence interval: 1,090.59 – 1,107.13

10184 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,092.63 rating

Confidence interval: 1,084.02 – 1,101.24

9591 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,070.70 rating

Confidence interval: 1,063.09 – 1,078.31

12635 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

926.88 rating

Confidence interval: 893.47 – 960.30

341 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

963.98 rating

Confidence interval: 924.51 – 1,003.45

243 votes

Oct 28, 2026

Category: korean
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,061.75 rating

Confidence interval: 1,049.13 – 1,074.37

2813 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,091.22 rating

Confidence interval: 1,080.95 – 1,101.50

4740 votes

Oct 28, 2026

Category: math
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,088.03 rating

Confidence interval: 1,077.58 – 1,098.47

5533 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,045.11 rating

Confidence interval: 1,037.32 – 1,052.91

12447 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,115.20 rating

Confidence interval: 1,109.76 – 1,120.65

38492 votes

Oct 28, 2026

Category: overall
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,082.68 rating

Confidence interval: 1,068.20 – 1,097.15

1981 votes

Oct 28, 2026

Category: russian
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

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

Llama-2-70b-chat

1,142.49 rating

Confidence interval: 1,113.38 – 1,171.60

506 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: Llama 2 70B Chat
Evaluation type: human preference

Arena text preference / chinesehuman preference · 2026-10-15 · rating

Llama-2-70b-chat

994.63 rating

Confidence interval: 981.25 – 1,008.00

3619 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,078.59 rating

Confidence interval: 1,068.78 – 1,088.40

5717 votes

Oct 15, 2026

Category: coding
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,074.74 rating

Confidence interval: 1,064.69 – 1,084.79

6376 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,155.92 rating

Confidence interval: 1,149.50 – 1,162.35

26045 votes

Oct 15, 2026

Category: english
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

Llama-2-70b-chat

955.09 rating

Confidence interval: 947.04 – 963.14

25547 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,039.10 rating

Confidence interval: 1,021.92 – 1,056.28

1356 votes

Oct 15, 2026

Category: expert
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,090.06 rating

Confidence interval: 1,062.03 – 1,118.09

679 votes

Oct 15, 2026

Category: french
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / germanhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,041.26 rating

Confidence interval: 1,018.88 – 1,063.65

878 votes

Oct 15, 2026

Category: german
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,072.89 rating

Confidence interval: 1,064.76 – 1,081.01

9650 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,091.15 rating

Confidence interval: 1,081.98 – 1,100.32

6989 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,082.80 rating

Confidence interval: 1,070.84 – 1,094.75

3543 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,064.43 rating

Confidence interval: 1,054.53 – 1,074.33

6070 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,113.79 rating

Confidence interval: 1,098.32 – 1,129.25

2073 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,119.73 rating

Confidence interval: 1,109.77 – 1,129.69

6638 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,089.42 rating

Confidence interval: 1,078.41 – 1,100.43

4052 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,093.17 rating

Confidence interval: 1,077.75 – 1,108.59

2169 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,098.94 rating

Confidence interval: 1,090.67 – 1,107.22

10184 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,092.69 rating

Confidence interval: 1,084.08 – 1,101.31

9591 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,070.71 rating

Confidence interval: 1,063.10 – 1,078.33

12635 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-15 · rating

Llama-2-70b-chat

926.71 rating

Confidence interval: 893.34 – 960.09

341 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

964.02 rating

Confidence interval: 924.56 – 1,003.48

243 votes

Oct 15, 2026

Category: korean
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,061.79 rating

Confidence interval: 1,049.16 – 1,074.42

2813 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,091.18 rating

Confidence interval: 1,080.91 – 1,101.46

4740 votes

Oct 15, 2026

Category: math
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,088.02 rating

Confidence interval: 1,077.58 – 1,098.47

5533 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,045.20 rating

Confidence interval: 1,037.41 – 1,053.00

12447 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,115.19 rating

Confidence interval: 1,109.74 – 1,120.64

38492 votes

Oct 15, 2026

Category: overall
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,082.83 rating

Confidence interval: 1,068.35 – 1,097.31

1981 votes

Oct 15, 2026

Category: russian
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-15 · rating

Llama-2-70b-chat

1,142.41 rating

Confidence interval: 1,113.27 – 1,171.56

506 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: llama-2-70b-chat
Evaluation type: human preference

Run it locally

Inference · memory estimate

Llama-2-70b-chat

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Open weights: Yes
Model license: llama2

Input modalities

text

Output modalities

text