Loading...

Loading...

Buy Me A Coffee

Contribute Sources

Social

Support

© 2025 AI StatsIf you run into any issues or notice any data errors, please visit our GitHub and report an issue.Last updated: just now

Home Comparisons Providers Models Benchmarks Prices

Thematic Generalisation

36

Total Models

1.99

Average Score

1.70 - 2.60

Score Range

1

Max Score Achievable

Top 10 Model Performance

Top 10 of 36

Lower is better

Models Using This Benchmark(36)Lower is better

OpenAI

OpenAI(9 models)

openai

1.80

openai

1.82

openai

1.82

openai

1.83

openai

1.85

openai

1.93

openai

1.95

openai

1.96

openai

2.30

Anthropic

Anthropic(5 models)

anthropic

1.70

Claude 3.7 Sonnet

anthropic

1.88

Claude Sonnet 4

anthropic

1.89

Claude 3.5 Sonnet

anthropic

1.93

Claude 3.5 Haiku

anthropic

2.25

Google

Google(5 models)

Gemini 2.5 Pro Experimental

google

1.74

Gemini 2.5 Pro Preview

google

1.75

Gemini 2.5 Pro Preview

google

1.79

google

2.21

google

2.60

xAI

xAI(4 models)

x-ai

1.88

Grok 3 Mini Beta

x-ai

1.90

x-ai

2.07

x-ai

2.21

DeepSeek

DeepSeek(3 models)

deepseek

1.74

DeepSeek-V3 0324

deepseek

1.95

deepseek

2.03

Meta

Meta(3 models)

Llama 4 Maverick

meta

2.04

Llama 3.1 405B (base)

meta

2.08

Llama 3.3 70B Instruct

meta

2.12

Qwen

Qwen(3 models)

Qwen3 235B A22B

qwen

1.90

qwen

2.09

Qwen2.5 72B Instruct

qwen

2.21

Amazon

Amazon(1 model)

amazon

2.11

Microsoft

Microsoft(1 model)

microsoft

2.10

Mistral

Mistral(1 model)

Mistral Large 2

mistral

2.11

Moonshot

Moonshot(1 model)

Kimi K2 Instruct

moonshotai

1.94