Llama-3.3-70B-Instruct
Stored tensors can include buffers and repeated shared weights. A verified unique parameter count is unavailable. Checkpoint size is not the memory required to run the model.
GeekRadar / The comparison desk
COMPARE.
AI modelsLlama-3.3-70B-Instruct
AI model Released
- Context window
- 128,000 tokens
- Maximum output
- 4,096 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.
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-3.3-70B-Instruct
1,216.95 rating
Confidence interval: 1,206.44 – 1,227.45
3676 votes
- Category
- chinese
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,267.82 rating
Confidence interval: 1,261.29 – 1,274.36
8735 votes
- Category
- coding
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,250.12 rating
Confidence interval: 1,243.13 – 1,257.12
8045 votes
- Category
- creative_writing
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,303.43 rating
Confidence interval: 1,299.11 – 1,307.75
29254 votes
- Category
- english
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,199.25 rating
Confidence interval: 1,194.25 – 1,204.25
36733 votes
- Category
- exclude_ties
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / experthuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,225.21 rating
Confidence interval: 1,214.09 – 1,236.34
2917 votes
- Category
- expert
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,281.56 rating
Confidence interval: 1,256.46 – 1,306.65
574 votes
- Category
- french
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,250.74 rating
Confidence interval: 1,233.73 – 1,267.75
1223 votes
- Category
- german
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,257.07 rating
Confidence interval: 1,252.10 – 1,262.03
16962 votes
- Category
- hard_prompts
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,276.75 rating
Confidence interval: 1,270.49 – 1,283.02
9545 votes
- Category
- hard_prompts_english
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,242.82 rating
Confidence interval: 1,235.25 – 1,250.39
6627 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,240.60 rating
Confidence interval: 1,234.31 – 1,246.89
9713 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,284.24 rating
Confidence interval: 1,273.39 – 1,295.09
3154 votes
- Category
- industry_legal_and_government
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,284.01 rating
Confidence interval: 1,277.51 – 1,290.50
9353 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,267.05 rating
Confidence interval: 1,258.93 – 1,275.18
5237 votes
- Category
- industry_mathematical
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,275.12 rating
Confidence interval: 1,263.16 – 1,287.07
2730 votes
- Category
- industry_medicine_and_healthcare
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,272.82 rating
Confidence interval: 1,267.46 – 1,278.17
14851 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,254.93 rating
Confidence interval: 1,249.34 – 1,260.52
13804 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,241.67 rating
Confidence interval: 1,236.65 – 1,246.68
18780 votes
- Category
- instruction_following
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,150.47 rating
Confidence interval: 1,131.88 – 1,169.06
1170 votes
- Category
- japanese
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,142.44 rating
Confidence interval: 1,120.54 – 1,164.35
856 votes
- Category
- korean
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,255.29 rating
Confidence interval: 1,248.65 – 1,261.93
8327 votes
- Category
- longer_query
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,267.13 rating
Confidence interval: 1,259.38 – 1,274.87
5778 votes
- Category
- math
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,279.48 rating
Confidence interval: 1,272.84 – 1,286.13
9058 votes
- Category
- multi_turn
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,236.17 rating
Confidence interval: 1,231.74 – 1,240.59
25237 votes
- Category
- non_english
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,274.23 rating
Confidence interval: 1,270.75 – 1,277.72
54493 votes
- Category
- overall
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,259.07 rating
Confidence interval: 1,242.32 – 1,275.82
1454 votes
- Category
- polish
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,251.51 rating
Confidence interval: 1,243.56 – 1,259.45
5712 votes
- Category
- russian
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating
Llama-3.3-70B-Instruct
1,269.55 rating
Confidence interval: 1,245.52 – 1,293.58
636 votes
- Category
- spanish
- Evaluated configuration
- Llama 3.3 70B Instruct
- Evaluation type
- human preference
Arena text preference / chinesehuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,217.02 rating
Confidence interval: 1,206.52 – 1,227.53
3676 votes
- Category
- chinese
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,267.93 rating
Confidence interval: 1,261.39 – 1,274.46
8735 votes
- Category
- coding
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,250.23 rating
Confidence interval: 1,243.24 – 1,257.23
8046 votes
- Category
- creative_writing
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,303.35 rating
Confidence interval: 1,299.03 – 1,307.67
29253 votes
- Category
- english
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,199.24 rating
Confidence interval: 1,194.24 – 1,204.24
36733 votes
- Category
- exclude_ties
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / experthuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,225.25 rating
Confidence interval: 1,214.13 – 1,236.38
2917 votes
- Category
- expert
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,281.48 rating
Confidence interval: 1,256.43 – 1,306.53
575 votes
- Category
- french
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,250.74 rating
Confidence interval: 1,233.76 – 1,267.71
1223 votes
- Category
- german
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,257.02 rating
Confidence interval: 1,252.06 – 1,261.98
16961 votes
- Category
- hard_prompts
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,276.59 rating
Confidence interval: 1,270.33 – 1,282.85
9544 votes
- Category
- hard_prompts_english
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,242.85 rating
Confidence interval: 1,235.28 – 1,250.41
6629 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,240.65 rating
Confidence interval: 1,234.36 – 1,246.94
9715 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,283.94 rating
Confidence interval: 1,273.09 – 1,294.79
3153 votes
- Category
- industry_legal_and_government
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,284.00 rating
Confidence interval: 1,277.50 – 1,290.49
9352 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,267.06 rating
Confidence interval: 1,258.95 – 1,275.18
5237 votes
- Category
- industry_mathematical
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,275.24 rating
Confidence interval: 1,263.29 – 1,287.18
2731 votes
- Category
- industry_medicine_and_healthcare
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,272.92 rating
Confidence interval: 1,267.57 – 1,278.27
14851 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,254.98 rating
Confidence interval: 1,249.39 – 1,260.57
13808 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,241.66 rating
Confidence interval: 1,236.65 – 1,246.67
18783 votes
- Category
- instruction_following
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,150.22 rating
Confidence interval: 1,131.69 – 1,168.75
1171 votes
- Category
- japanese
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,142.59 rating
Confidence interval: 1,120.74 – 1,164.45
857 votes
- Category
- korean
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,255.35 rating
Confidence interval: 1,248.71 – 1,261.99
8330 votes
- Category
- longer_query
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,267.03 rating
Confidence interval: 1,259.29 – 1,274.77
5778 votes
- Category
- math
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,279.49 rating
Confidence interval: 1,272.85 – 1,286.13
9059 votes
- Category
- multi_turn
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,236.26 rating
Confidence interval: 1,231.84 – 1,240.68
25242 votes
- Category
- non_english
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,274.21 rating
Confidence interval: 1,270.72 – 1,277.69
54497 votes
- Category
- overall
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,259.17 rating
Confidence interval: 1,242.42 – 1,275.91
1454 votes
- Category
- polish
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,251.67 rating
Confidence interval: 1,243.72 – 1,259.61
5713 votes
- Category
- russian
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating
Llama-3.3-70B-Instruct
1,269.47 rating
Confidence interval: 1,245.45 – 1,293.50
636 votes
- Category
- spanish
- Evaluated configuration
- llama-3.3-70b-instruct
- Evaluation type
- human preference
Run it locally
API cost
Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.
Llama-3.3-70B-Instruct
Llama-3.3-70B-Instruct / Pricing tiers
| Provider / scope | Rate | Currency | Amount / unit |
|---|---|---|---|
| DeepInframeta-llama/Llama-3.3-70B-Instruct-TurboStandard | Input | USD | 0.1 / 1M tokens |
| Output | USD | 0.32 / 1M tokens | |
| SambaNova CloudMeta-Llama-3.3-70B-InstructStandard | Input | USD | 0.6 / 1M tokens |
| Output | USD | 1.2 / 1M tokens |
Capabilities
- Context window
- 128,000 tokens
- Maximum output
- 4,096 tokens
- Reasoning
- No
- Tool calling
- Yes
- Open weights
- Yes
- Knowledge cutoff
- 2023-12
- Checkpoint tensor payload
- 131.42 GiB
Input modalities
- text
Output modalities
- text
0 selected Compare 0