Lithium carbon dioxide (Li-CO2) batteries have attracted significant attention due to their dual functionality to store energy and simultaneously capture CO2 gas. Cathodes based on carbon nanomaterials, such as graphene and carbon nanotubes (CNTs), have shown high catalytic performance in Li-CO2 batteries. Optimising the cathode composition and structure using traditional experimental and computational techniques is expensive and time-consuming. Machine learning (ML) techniques can significantly facilitate optimisation processes and forecast cathode performance, including estimates of remaining cycle life (RCL) and capacity. BN-hG, COF-Ru@CNT, BNCNT, Ruh-NC@rGO, Rul-NC@rGO, and Runp-NC@rGO are the six carbon-based cathodes assessed in this study using a 1 M LiTFSI/TEGDME electrolyte at a current density of 1000 mA g-1 and a cutoff capacity of 1000 mAh g-1. To train various supervised ML regression models, such as Linear, Polynomial, K-Nearest Neighbours (KNN), Classification & Regression Trees (C&RT), Support Vector Machine (SVM), Multi-variate Adaptive Regression Splines (MARSpline), and Statistical Artificial Neural Network (SANN), cycle number and discharge voltage were trained as input features. The models that showed high accuracy (R2 > 0.95) and good fit to the training data are filtered. To evaluate the predictions, various factors affecting the cycle life are calculated, including changes in internal resistance, gravimetric energy density, energy efficiency, polarisation resistance, and the effect of voltages, and are supported by theoretical reasoning. Both Ruh-NC@rGO and COF-Ru@CNT demonstrated low polarisation, excellent charge-transfer kinetics, and stable Li2CO3 formation and decomposition, underscoring their potential as high-performance catalysts. This study demonstrates how to incorporate ML into the study of discharge-voltage behaviour in Li-CO2 batteries without optimising the data, and it can advance next-generation high-energy storage systems by estimating RCL without affecting performance beyond the electrochemical window.