The growing global demand for sustainable and high-efficiency energy sources has intensified research on alternative fuels for spark-ignition engines. This study experimentally investigates the performance and emission characteristics of the EF7 engine fueled with gasoline blends containing ethanol, methanol, and toluene at 5%, 10%, and 15% volumetric concentrations. An artificial neural network (ANN) model was concurrently developed to predict critical exhaust emissions including CO, CO₂, NOx, and HC based on engine operating parameters. Experiments were conducted at full load under constant speeds of 2000 and 3000 rpm, and the ANN model was trained and validated using the collected datasets. Results indicate that ethanol and methanol blends enhance brake power and torque up to a 10% blending ratio, primarily due to increased laminar flame speed, charge-cooling effects, and oxygen-enriched combustion chemistry. However, further increasing the alcohol fraction to 15% reduced engine output, attributed to the lower heating value and diminished volumetric energy density of the blends. Toluene addition provided stable power and torque across all blending ratios owing to its high octane number and knock resistance, though without notable performance gains. Emission analysis revealed that alcohol blends significantly decreased CO and HC emissions while slightly elevating CO₂ levels, confirming improved combustion efficiency. In contrast, toluene blends increased NOx and HC emissions, likely resulting from higher local combustion temperatures and incomplete aromatic oxidation. Overall, moderate alcohol blending (approximately 10%) achieved the optimal trade-off between engine performance and emission reduction. The integrated experimental–computational framework established in this study offers a robust methodology for optimizing blended fuels in spark-ignition engines.