Measuring LLM Personality: A Quantitative Comparison of GPT-5.2 and Opus 4.5 (www.lindr.io)

A recent study from Lindr has introduced a quantitative framework for measuring the personality of large language models (LLMs), specifically comparing GPT-5.2 and Claude Opus 4.5 across various contexts. Using 500 unique prompts spanning ten personality dimensions, the evaluation revealed that Claude exhibits higher levels of Openness and Curiosity, while GPT-5.2 excels in Conscientiousness. The effect sizes indicate these differences are statistically significant, with model identity explaining approximately 45% of the variance in personality scores, which highlights the importance of model selection in user-facing applications. This research is crucial for the AI/ML community as it paves the way for developing personalized AI interactions that align with user expectations. The findings suggest that personality traits vary not only by model but also by context, implying that LLMs should be tailored for specific applications, such as creative versus task-oriented tasks. The Lindr personality evaluation API now offers a systematic method for assessing these traits, encouraging developers to consider personality in their design processes and raising important questions about the stability and adaptability of AI personalities.