GeekRadar / The comparison desk

COMPARE.

AI models

allenai

Llama-3.1-Tulu-3-8B

AI model

Context window: 131,072 tokens
Total parameters: 8.03 B
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-3.1-Tulu-3-8B 1193.3 Llama 3.1 Tulu 3 8B · 95% interval 1182.8–1203.8 · 2,896 votes · 2026-11-06

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 · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,175.70 rating

Confidence interval: 1,139.87 – 1,211.53

238 votes

Nov 6, 2026

Category: chinese
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,183.17 rating

Confidence interval: 1,158.92 – 1,207.43

476 votes

Nov 6, 2026

Category: coding
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,181.16 rating

Confidence interval: 1,156.11 – 1,206.21

472 votes

Nov 6, 2026

Category: creative_writing
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,214.86 rating

Confidence interval: 1,201.26 – 1,228.47

1633 votes

Nov 6, 2026

Category: english
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,059.80 rating

Confidence interval: 1,040.91 – 1,078.69

1763 votes

Nov 6, 2026

Category: exclude_ties
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,174.04 rating

Confidence interval: 1,154.51 – 1,193.58

728 votes

Nov 6, 2026

Category: hard_prompts
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,177.73 rating

Confidence interval: 1,152.56 – 1,202.90

452 votes

Nov 6, 2026

Category: hard_prompts_english
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,196.30 rating

Confidence interval: 1,164.75 – 1,227.85

285 votes

Nov 6, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,160.79 rating

Confidence interval: 1,138.14 – 1,183.44

563 votes

Nov 6, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,193.30 rating

Confidence interval: 1,168.95 – 1,217.65

516 votes

Nov 6, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,192.75 rating

Confidence interval: 1,163.59 – 1,221.91

304 votes

Nov 6, 2026

Category: industry_mathematical
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,191.38 rating

Confidence interval: 1,171.79 – 1,210.97

761 votes

Nov 6, 2026

Category: industry_software_and_it_services
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,177.49 rating

Confidence interval: 1,158.19 – 1,196.80

816 votes

Nov 6, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,174.05 rating

Confidence interval: 1,158.64 – 1,189.46

1172 votes

Nov 6, 2026

Category: instruction_following
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,180.43 rating

Confidence interval: 1,155.07 – 1,205.79

454 votes

Nov 6, 2026

Category: longer_query
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,194.74 rating

Confidence interval: 1,169.39 – 1,220.09

363 votes

Nov 6, 2026

Category: math
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,153.51 rating

Confidence interval: 1,128.41 – 1,178.61

520 votes

Nov 6, 2026

Category: multi_turn
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,168.83 rating

Confidence interval: 1,153.82 – 1,183.83

1263 votes

Nov 6, 2026

Category: non_english
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,193.27 rating

Confidence interval: 1,182.78 – 1,203.76

2896 votes

Nov 6, 2026

Category: overall
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-11-06 · rating

Llama-3.1-Tulu-3-8B

1,193.05 rating

Confidence interval: 1,168.54 – 1,217.57

418 votes

Nov 6, 2026

Category: russian
Evaluated configuration: Llama 3.1 Tulu 3 8B
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,175.61 rating

Confidence interval: 1,139.76 – 1,211.46

238 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,183.25 rating

Confidence interval: 1,158.98 – 1,207.52

476 votes

Oct 15, 2026

Category: coding
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,181.33 rating

Confidence interval: 1,156.26 – 1,206.40

472 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,214.79 rating

Confidence interval: 1,201.18 – 1,228.39

1633 votes

Oct 15, 2026

Category: english
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,059.86 rating

Confidence interval: 1,040.96 – 1,078.76

1763 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,174.13 rating

Confidence interval: 1,154.58 – 1,193.68

728 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,177.68 rating

Confidence interval: 1,152.50 – 1,202.86

452 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,196.44 rating

Confidence interval: 1,164.87 – 1,228.02

285 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,160.83 rating

Confidence interval: 1,138.18 – 1,183.49

563 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,193.31 rating

Confidence interval: 1,168.96 – 1,217.67

516 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,192.61 rating

Confidence interval: 1,163.44 – 1,221.78

304 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,191.49 rating

Confidence interval: 1,171.89 – 1,211.09

761 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,177.58 rating

Confidence interval: 1,158.27 – 1,196.89

816 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,174.05 rating

Confidence interval: 1,158.63 – 1,189.46

1172 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,180.46 rating

Confidence interval: 1,155.08 – 1,205.84

454 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,194.69 rating

Confidence interval: 1,169.34 – 1,220.04

363 votes

Oct 15, 2026

Category: math
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,153.58 rating

Confidence interval: 1,128.47 – 1,178.69

520 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,168.97 rating

Confidence interval: 1,153.96 – 1,183.97

1263 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,193.27 rating

Confidence interval: 1,182.77 – 1,203.77

2896 votes

Oct 15, 2026

Category: overall
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

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

Llama-3.1-Tulu-3-8B

1,193.11 rating

Confidence interval: 1,168.58 – 1,217.64

418 votes

Oct 15, 2026

Category: russian
Evaluated configuration: llama-3.1-tulu-3-8b
Evaluation type: human preference

Run it locally

Inference · memory estimate

Llama-3.1-Tulu-3-8B

8.03 B total parameters · llama · BF16

Published checkpoint tensor payload: 15.0 GiB. This includes stored copies and buffers; it is separate from the runtime estimates below.

Weight precision Total context Check a GPU Check a Local AI system

Estimated accelerator memory5.6 GiB

Weights: 4.1 GiB
KV cache: 0.5 GiB
Runtime allowance: 1.0 GiB

One user, full model in accelerator memory.

GPUs with memory headroom

Local AI systems with memory headroom

NVIDIA Jetson Orin Nano Super Developer Kit (8 GB) Planning budget: 6.0 GiB
NVIDIA Jetson AGX Orin Developer Kit (64 GB) Planning budget: 48.0 GiB
AMD Ryzen AI Halo (Ryzen AI Max+ 395, 128 GB) Planning budget: 96.0 GiB
ASUS Ascent GX10 (128 GB) Planning budget: 96.0 GiB
Apple Mac Studio (M5 Max 18-core CPU, 40-core GPU, 128 GB) Planning budget: 96.0 GiB
Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB) Planning budget: 96.0 GiB

Shared-memory systems use a planning budget of 75% of installed memory, leaving 25% for the OS and other allocations. This is an assumption, not a measured GPU allocation. Firmware and driver limits may be lower and may require configuration. Memory matches do not verify a runtime, quantized model artifact or inference speed.

Includes FP16 KV cache, prompt plus generated tokens, quantization metadata and a runtime allowance. GPU matches reserve 10% of VRAM. Memory fit is not a speed or software-compatibility guarantee; quantized weights and runtime support must be available.

Capabilities

Context window: 131,072 tokens
Open weights: Yes
Model license: llama3.1
Total parameters: 8.03 B
Checkpoint tensor payload: 14.96 GiB

Input modalities

text

Output modalities

text

Processing & platform

Architecture: llama