This study explores the effectiveness of using GPT, a Large Language Model (LLM), for Named Entity Recognition (NER) in Spanish electronic health records (EHRs) related to breast cancer. The research compares the performance of GPT with BERT, a traditional NER model, using fine-tuning techniques and prompting methods. The dataset consists of Spanish EHRs, and the evaluation metrics include precision, recall, and F-score. The study finds that both BERT and GPT achieve comparable levels of precision, with GPT slightly outperforming BERT in recall and overall F-score. GPT's ability to handle complex medical terminologies and contextual nuances with minimal data annotation highlights its potential in medical data processing. The research also discusses the advantages of GPT, such as reduced computational resources and adaptability to new domains, making it a compelling option for diverse applications in healthcare.This study explores the effectiveness of using GPT, a Large Language Model (LLM), for Named Entity Recognition (NER) in Spanish electronic health records (EHRs) related to breast cancer. The research compares the performance of GPT with BERT, a traditional NER model, using fine-tuning techniques and prompting methods. The dataset consists of Spanish EHRs, and the evaluation metrics include precision, recall, and F-score. The study finds that both BERT and GPT achieve comparable levels of precision, with GPT slightly outperforming BERT in recall and overall F-score. GPT's ability to handle complex medical terminologies and contextual nuances with minimal data annotation highlights its potential in medical data processing. The research also discusses the advantages of GPT, such as reduced computational resources and adaptability to new domains, making it a compelling option for diverse applications in healthcare.