OpenAI's o3 model has achieved an IQ score of 135 in a rigorous IQ test, surpassing the human average of 90-110.
Other top-performing AI models include Anthropic's Claude-4 Sonnet at 127 and Google's Gemini 2.0 Flash at 126.
Language-based AI models outperformed vision-enabled systems in the test, with models like GPT-4o with vision scoring 63.
The results highlight the advancement of AI in language-based reasoning but also underscore the challenge in developing genuine multimodal reasoning abilities.