Objective: This study explores the determination of audit fees (AF) using Support Vector Regression (SVR) among companies listed on the Iranian stock market from 2017 to 2021. It investigates the relationship between financial variables like financial leverage (DA), current assets ratio (CA), quick ratio (QUICK), ASSETS, current ratio to current liabilities (CR), and long-term debt (DE), with AF as the target. Methodology: Data from 60 listed companies during this period, totaling 279 year-observations, are employed. SVR models are trained on this dataset using Google Colab. Results: The SVR model achieves a 90.5% R2 value and a 3.7 Mean Squared Error (MSE) on training data, indicating high explained variance and reasonable error levels. However, on new data, the model's performance diminishes, with an R2 of 67% and an MSE of 8.1, implying reduced accuracy and intermediate predictive accuracy. Innovation: This study advances the understanding of AF determination using SVR, highlighting the importance of considering various financial variables. تفاصيل المقالة