The growing global adoption of electric vehicles (EVs) emphasizes the pressing need for a comprehensive understanding of thermal runaway in lithium-ion batteries. Prevention of the onset of thermal runaway and its subsequent propagation throughout the entire battery pack is one of the pressing challenges of lithium-ion batteries. In addition to generating excess heat, thermal runaway of batteries also releases hazardous flammable gases, posing risks of external combustion and fires. Most existing thermal runaway models in literature primarily focus on predicting heat release or the total amount of vent gas. In this study, we present a model capable of predicting both heat release and the transient composition of emitted gases, including CO, H2, CO2, and hydrocarbons, during thermal runaway events. We calibrated the model using experimental data obtained from an 18650 cell from the literature, ensuring the accuracy of reaction parameters. We employ this developed model to investigate how different state-of-charge (SOC) levels (25%, 50%, 75%, and 100%) impact thermal runaway events and subsequent gas composition. Our analysis of three major input parameters: pre-exponent multiplier, activation energy, and specific heat release, across the SOC levels, revealed similar reaction rates for SOC levels between 50% and 100%, except for the anode, with significant difference for 25% SOC parameters, resulting in lower cell temperatures during thermal runaway.