The driving concept behind TradeGPT is rooted in the findings from the paper "Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models". To make it easier for humans to leverage AI in trading, I've tailored the algorithm to focus on the top 15 stocks by trading volume. Rather than spreading our investments across many stocks based on average sentiment scores, we've refined the strategy to concentrate on the top three stocks from this select group. Each day, I collect the latest news headlines for each of these 15 stocks and analyze their potential impact on stock prices with the help of ChatGPT. Each headline is scored on a scale from -1 to 1, indicating negative or positive impact respectively. I then calculate the average of these scores for each stock, rank these averages across all 15 stocks, and select the top three for trading. TradeGPT performs this analysis precisely at midnight each trading day. We compute the daily return by averaging the ratio of the closing to the opening price of each of the top three selected stocks. This streamlined approach not only enhances our trading strategy but also provides clear, actionable insights into stock market dynamics.