<p>The practice of raising guinea pigs for meat production has garnered significant global attention. This meat offers a variety of benefits for producers and consumers, including the enhancement of economic income for rural families and the broadening of the gastronomic landscape in certain nations. On the other hand, the management of guinea pig production systems necessitates information to support decisions regarding selection criteria, standardization of slaughter batches, and prediction of meat yield with the aim of meeting market demand, among other considerations. The measurements obtained from the pre- and post-slaughter processes have the potential to predict the carcass tissue composition, which can be used under experimental conditions. The aim of this study was to compare conventional prediction models to predict guinea pig carcass composition, using or not a lower commercial value cut, and to propose an artificial neural network as a method to improve the accuracy. A total of 130 male and female guinea pigs of varying ages were slaughtered, and the quality traits of their carcasses were meticulously documented. The primary findings showed good prediction models derived from multiple linear regression analyses, which incorporated animal-specific weights and linear carcass measurements. The incorporation of neck traits data into the LRM led to an enhancement in the accuracy of the model. Nevertheless, the most accurate results were obtained with methods based on machine learning. The proposed prediction models have the potential to benefit guinea pig meat farmers and research institutions by identifying optimal practices for predicting guinea pig carcass composition. This can be achieved through the utilization of traditional multiple linear regression or machine learning methods, leading to the enhancement of carcass component or tissue composition measurement techniques that are both time- and cost-efficient.</p>