Side-by-Side Uncensored LLM Comparison Matrix

Filter and select up to 4 open-weight uncensored models to compare UGI compliance, MMLU-Pro reasoning retention, EQ-Bench nuance, and quantized GPU VRAM footprints side-by-side.

1-Click Comparison Presets0/4 Active

Select Models to Compare (0/4 Selected)

· 10 matching models

VRAM:

#1 · 405BScore 93.5

Hermes-3-Llama-3.1-405B

Llama 3.1 · 228GB Q4

96.8% compl.Compare

#2 · 32.8BScore 90.8

DeepSeek-R1-Distill-Qwen-32B-abliterated

Qwen 2.5 · 20GB Q4

97.4% compl.Compare

#3 · 70.6BScore 90.6

Llama-3.3-70B-Instruct-abliterated

Llama 3.3 · 41GB Q4

97.6% compl.Compare

#4 · 72.7BScore 90.4

Qwen2.5-72B-Instruct-abliterated

Qwen 2.5 · 42GB Q4

98.2% compl.Compare

#5 · 32.5BScore 90.0

Qwen2.5-Coder-32B-Instruct-abliterated

Qwen 2.5 Coder · 20GB Q4

98.5% compl.Compare

#6 · 24BScore 88.8

Dolphin3.0-R1-Mistral-24B

Mistral Small 3 · 15GB Q4

99.1% compl.Compare

#7 · 72.7BScore 88.5

EVA-Qwen2.5-72B-v0.2

Qwen 2.5 · 42GB Q4

98.9% compl.Compare

#8 · 72.7BScore 88.2

magnum-v4-72b

Qwen 2.5 · 42GB Q4

99.4% compl.Compare