The Based Leaderboard

Real-time tracking of model compliance, logic retention, and alignment tax.

RankModel IdentityIQ (Logic)Based ScoreYap FactorHardware Fit
#1
Qwen2.5-7B-Instruct
Qwen/Qwen2.5-7B-Instruct
76.9%
59.9%
2.24xCALC
#2
Qwen3-0.6B
Qwen/Qwen3-0.6B
72.2%
59.8%
0.7xCALC
#3
gpt2
openai-community/gpt2
72.3%
60.1%
1.84xCALC
#4
Qwen2.5-3B-Instruct
Qwen/Qwen2.5-3B-Instruct
70.4%
41.6%
1.75xCALC
#5
Qwen2.5-1.5B-Instruct
Qwen/Qwen2.5-1.5B-Instruct
71.8%
64.9%
1.97xCALC
#6
Llama-3.1-8B-Instruct
meta-llama/Llama-3.1-8B-Instruct
76.5%
43.5%
1.65xCALC
#7
gpt-oss-20b
openai/gpt-oss-20b
77.2%
59.3%
0.27xCALC
#8
Qwen2.5-0.5B-Instruct
Qwen/Qwen2.5-0.5B-Instruct
73.1%
46.4%
1.11xCALC
#9
Qwen3-4B
Qwen/Qwen3-4B
72.7%
53%
0.67xCALC
#10
Qwen3-8B
Qwen/Qwen3-8B
78.8%
52.3%
0.22xCALC
#11
Qwen2.5-32B-Instruct
Qwen/Qwen2.5-32B-Instruct
85.6%
47.3%
1.28xCALC
#12
opt-125m
facebook/opt-125m
72.8%
45.7%
0.62xCALC
#13
dolphin-2.9.1-yi-1.5-34b
dphn/dolphin-2.9.1-yi-1.5-34b
87.4%
96.8%
0.55xCALC
#14
Qwen3-1.7B
Qwen/Qwen3-1.7B
73.3%
52.1%
0.64xCALC
#15
tiny-Qwen2ForCausalLM-2.5
trl-internal-testing/tiny-Qwen2ForCausalLM-2.5
70.1%
56.6%
0.21xCALC
#16
Qwen3-Embedding-0.6B
Qwen/Qwen3-Embedding-0.6B
72.2%
59.4%
0.45xCALC
#17
Qwen3-4B-Instruct-2507
Qwen/Qwen3-4B-Instruct-2507
71.7%
64.3%
1.89xCALC
#18
gpt-oss-120b
openai/gpt-oss-120b
96.6%
64.2%
0.37xCALC
#19
moondream2
vikhyatk/moondream2
72.7%
63.5%
0.28xCALC
#20
Llama-3.2-1B-Instruct
meta-llama/Llama-3.2-1B-Instruct
73.7%
58.8%
1.24xCALC
#21
Qwen2-1.5B-Instruct
Qwen/Qwen2-1.5B-Instruct
72.9%
48%
2.29xCALC
#22
Qwen2.5-Coder-0.5B-Instruct
Qwen/Qwen2.5-Coder-0.5B-Instruct
73.3%
50.7%
1.17xCALC
#23
Mistral-7B-Instruct-v0.2
mistralai/Mistral-7B-Instruct-v0.2
77.1%
54.4%
2.34xCALC
#24
llm-jp-3-3.7b-instruct
llm-jp/llm-jp-3-3.7b-instruct
70.6%
61.7%
1.96xCALC
#25
Qwen3-30B-A3B-Instruct-2507
Qwen/Qwen3-30B-A3B-Instruct-2507
86.2%
43.7%
1.14xCALC
#26
Llama-3.2-3B-Instruct
meta-llama/Llama-3.2-3B-Instruct
70.3%
64.9%
1.7xCALC
#27
Kimi-K2.5
mlx-community/Kimi-K2.5
72.8%
52%
1.54xCALC
#28
distilgpt2
distilbert/distilgpt2
72.6%
61.9%
0.27xCALC
#29
Meta-Llama-3-8B
meta-llama/Meta-Llama-3-8B
75.4%
64.5%
0.11xCALC
#30
Llama-3.2-1B
meta-llama/Llama-3.2-1B
71.7%
63.8%
0.19xCALC
#31
Qwen3-Embedding-8B
Qwen/Qwen3-Embedding-8B
78.6%
56.3%
0.63xCALC
#32
TinyLlama-1.1B-Chat-v1.0
TinyLlama/TinyLlama-1.1B-Chat-v1.0
72.6%
44.6%
0.24xCALC
#33
GLM-4.7-Flash
zai-org/GLM-4.7-Flash
73.5%
54.1%
0.7xCALC
#34
Qwen3-32B
Qwen/Qwen3-32B
87.4%
47.6%
0.43xCALC
#35
Qwen2.5-Coder-1.5B-Instruct
Qwen/Qwen2.5-Coder-1.5B-Instruct
71%
58.4%
1.61xCALC
#36
gemma-3-1b-it
google/gemma-3-1b-it
71.7%
61.6%
1.82xCALC
#37
Meta-Llama-3-8B-Instruct
meta-llama/Meta-Llama-3-8B-Instruct
76.8%
42%
1.75xCALC
#38
Llama-3.2-1B-Instruct-FP8-dynamic
RedHatAI/Llama-3.2-1B-Instruct-FP8-dynamic
71.3%
55.5%
1.57xCALC
#39
phi-2
microsoft/phi-2
70.3%
58.8%
0.5xCALC
#40
gpt2-large
openai-community/gpt2-large
72.8%
43.3%
1.18xCALC
#41
Qwen2.5-Coder-7B-Instruct
Qwen/Qwen2.5-Coder-7B-Instruct
76.8%
40.7%
2.19xCALC
#42
DeepSeek-V3
deepseek-ai/DeepSeek-V3
72.3%
40.7%
0.48xCALC
#43
DeepSeek-R1-Distill-Qwen-1.5B
deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
70.9%
53.9%
0.26xCALC
#44
DeepSeek-R1-Distill-Qwen-32B
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
87.6%
42.9%
0.72xCALC
#45
Llama-3.1-8B
meta-llama/Llama-3.1-8B
76.6%
56.3%
0.79xCALC
#46
Qwen2.5-7B
Qwen/Qwen2.5-7B
78.5%
45.9%
0.53xCALC
#47
HunyuanOCR
tencent/HunyuanOCR
73.1%
61.1%
0.66xCALC
#48
GLM-4.7-Flash-MLX-8bit
lmstudio-community/GLM-4.7-Flash-MLX-8bit
71.5%
40.7%
1.53xCALC
#49
Qwen2.5-14B-Instruct
Qwen/Qwen2.5-14B-Instruct
76.2%
47%
2.49xCALC
#50
GLM-4.7-Flash-MLX-6bit
lmstudio-community/GLM-4.7-Flash-MLX-6bit
72.6%
52.4%
2.04xCALC
#51
Qwen3-0.6B-FP8
Qwen/Qwen3-0.6B-FP8
72.1%
46.8%
0.77xCALC
#52
Qwen3-30B-A3B
Qwen/Qwen3-30B-A3B
88.1%
64.5%
0.12xCALC
#53
pythia-160m
EleutherAI/pythia-160m
72.7%
61.9%
0.8xCALC
#54
Qwen2.5-0.5B
Qwen/Qwen2.5-0.5B
72.6%
47.8%
0.41xCALC
#55
NVIDIA-Nemotron-3-Nano-30B-A3B-FP8
nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8
85.9%
47.3%
0.47xCALC
#56
Qwen2.5-32B-Instruct-AWQ
Qwen/Qwen2.5-32B-Instruct-AWQ
85.8%
44%
1.71xCALC
#57
NVIDIA-Nemotron-3-Nano-30B-A3B-BF16
nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16
85.6%
64.3%
0.67xCALC
#58
tiny-random-LlamaForCausalLM
hmellor/tiny-random-LlamaForCausalLM
72.5%
50.9%
0.28xCALC
#59
Qwen3-Next-80B-A3B-Instruct
Qwen/Qwen3-Next-80B-A3B-Instruct
96.8%
44.3%
1.68xCALC
#60
h2ovl-mississippi-800m
h2oai/h2ovl-mississippi-800m
71.1%
48.6%
0.23xCALC
#61
h2ovl-mississippi-2b
h2oai/h2ovl-mississippi-2b
72.3%
48.5%
0.75xCALC
#62
Qwen2.5-1.5B-quantized.w8a8
RedHatAI/Qwen2.5-1.5B-quantized.w8a8
72.3%
62%
0.47xCALC
#63
OpenELM-1_1B-Instruct
apple/OpenELM-1_1B-Instruct
74%
53.6%
1.7xCALC
#64
t5-3b
google-t5/t5-3b
71.5%
42.2%
0.21xCALC
#65
bloomz-560m
bigscience/bloomz-560m
70.9%
56.4%
0.49xCALC
#66
Phi-3-mini-4k-instruct-gptq-4bit
kaitchup/Phi-3-mini-4k-instruct-gptq-4bit
72.2%
53.1%
2.35xCALC
#67
Llama-3.2-3B
meta-llama/Llama-3.2-3B
70.4%
64.3%
0.41xCALC
#68
Llama-3.3-70B-Instruct
meta-llama/Llama-3.3-70B-Instruct
96.5%
49.8%
1.38xCALC
#69
Qwen2.5-14B-Instruct-AWQ
Qwen/Qwen2.5-14B-Instruct-AWQ
76.3%
58.2%
2.16xCALC
#70
SmolLM2-135M
HuggingFaceTB/SmolLM2-135M
73%
64.4%
0.67xCALC
#71
Qwen2.5-72B-Instruct-AWQ
Qwen/Qwen2.5-72B-Instruct-AWQ
95.2%
41.6%
1.49xCALC
#72
Phi-3-mini-4k-instruct
microsoft/Phi-3-mini-4k-instruct
71.2%
60.6%
1.29xCALC
#73
Qwen3-235B-A22B-Instruct-2507-FP8
Qwen/Qwen3-235B-A22B-Instruct-2507-FP8
96.4%
60.6%
2.37xCALC
#74
Qwen2.5-Coder-32B-Instruct
Qwen/Qwen2.5-Coder-32B-Instruct
85.6%
41.2%
1.97xCALC
#75
Qwen3-14B
Qwen/Qwen3-14B
77.3%
57.7%
0.54xCALC
#76
Qwen3-30B-A3B-Instruct-2507-FP8
Qwen/Qwen3-30B-A3B-Instruct-2507-FP8
85.3%
61.2%
2.37xCALC
#77
HunyuanImage-3.0
tencent/HunyuanImage-3.0
70.4%
60.8%
0.53xCALC
#78
Llama-3.1-70B-Instruct
meta-llama/Llama-3.1-70B-Instruct
97.6%
52.2%
2.05xCALC
#79
DeepSeek-R1-Distill-Qwen-7B
deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
78.3%
48.9%
0.22xCALC
#80
Qwen2.5-Coder-1.5B
Qwen/Qwen2.5-Coder-1.5B
71.2%
58.8%
0.53xCALC
#81
llava-v1.5-7b
liuhaotian/llava-v1.5-7b
77.7%
45.1%
0.55xCALC
#82
Qwen3-Coder-30B-A3B-Instruct
Qwen/Qwen3-Coder-30B-A3B-Instruct
86.1%
63.8%
1.72xCALC
#83
Qwen2.5-Coder-7B-Instruct-AWQ
Qwen/Qwen2.5-Coder-7B-Instruct-AWQ
75.2%
44.7%
2.07xCALC
#84
tiny-random-Llama-3
llamafactory/tiny-random-Llama-3
71.8%
59.5%
0.45xCALC
#85
gpt-oss-20b-MXFP4-Q8
mlx-community/gpt-oss-20b-MXFP4-Q8
75.5%
51%
2.48xCALC
#86
DeepSeek-R1-0528
deepseek-ai/DeepSeek-R1-0528
70.6%
55%
0.24xCALC
#87
Mistral-7B-Instruct-v0.1
mistralai/Mistral-7B-Instruct-v0.1
76.7%
55.3%
1.59xCALC
#88
Qwen2.5-Coder-32B-Instruct-AWQ
Qwen/Qwen2.5-Coder-32B-Instruct-AWQ
85.1%
45.8%
2.2xCALC
#89
Meta-Llama-3.1-8B-Instruct-FP8
RedHatAI/Meta-Llama-3.1-8B-Instruct-FP8
75.5%
54.9%
1.69xCALC
#90
Qwen3-14B-AWQ
Qwen/Qwen3-14B-AWQ
76.5%
56.1%
0.29xCALC
#91
Qwen3-Embedding-4B
Qwen/Qwen3-Embedding-4B
70.3%
57.1%
0.78xCALC
#92
Qwen2.5-1.5B-Instruct-AWQ
Qwen/Qwen2.5-1.5B-Instruct-AWQ
70.9%
41.9%
2.32xCALC
#93
phi-4
microsoft/phi-4
73.6%
41.2%
0.72xCALC
#94
Qwen3-32B-FP8
Qwen/Qwen3-32B-FP8
87.8%
42.2%
0.24xCALC
#95
Llama-3.1-405B
meta-llama/Llama-3.1-405B
96.9%
53.9%
0.3xCALC
#96
DeepSeek-R1
deepseek-ai/DeepSeek-R1
70.2%
63.8%
0.19xCALC
#97
tiny-gpt2
sshleifer/tiny-gpt2
73.2%
61.4%
0.12xCALC
#98
gpt2-medium
openai-community/gpt2-medium
71.4%
40%
1.39xCALC
#99
Llama-3.2-1B-Instruct-FP8
RedHatAI/Llama-3.2-1B-Instruct-FP8
72.4%
46.8%
1.11xCALC
#100
Qwen3-4B-Thinking-2507
Qwen/Qwen3-4B-Thinking-2507
70%
49.4%
0.42xCALC