Loading...
Loading...
Confabulations
Twitter44
Total Models
19.16
Average Score
10.34 - 40.26
Score Range
1
Max Score Achievable
Top 10 Model Performance
Top 10 of 44
Lower is better
Models Using This Benchmark(44)Lower is better
OpenAI(14 models)
GPT-5
openai
10.34
o1
openai
11.74
o1 preview
openai
13.04
GPT-5 mini
openai
13.28
GPT-4.5
openai
13.64
o3 Pro
openai
14.22
o3
openai
14.38
GPT-4o
openai
15.34
gpt-oss-120b
openai
15.65
o4 Mini
openai
15.79
GPT-4o
openai
17.21
o3-mini
openai
18.43
o1 mini
openai
18.55
GPT-4o-mini
openai
37.21
Anthropic(8 models)
Claude Sonnet 4
anthropic
14.85
Claude Opus 4
anthropic
17.06
Claude Opus 4.1
anthropic
18.51
Claude 3.7 Sonnet
anthropic
19.76
Claude 3.5 Sonnet
anthropic
19.94
Claude 3 Opus
anthropic
22.70
Claude 3 Haiku
anthropic
34.21
Claude 3.5 Haiku
anthropic
36.74
Google(5 models)
Gemini 2.5 Pro Preview
google
10.62
10.80
Gemini 2.5 Pro Preview
google
12.38
Gemma 2 27B
google
27.24
Gemma 3 27B
google
40.26
Qwen(4 models)
xAI(4 models)
Meta(3 models)
DeepSeek(2 models)
R1
deepseek
14.56
DeepSeek-V3 0324
deepseek
26.15
Amazon(1 model)
Nova Pro 1.0
amazon
30.05
Microsoft(1 model)
Phi 4
microsoft
29.43
Mistral(1 model)
Mistral Large 2
mistral
21.40
Moonshot(1 model)
Kimi K2 Instruct
moonshotai
20.38