Purpose: To develop and validate a novel, hardware-free, smartphone-based video uroflowmetry system powered by Artificial Intelligence (AI-VUF), by comparing its diagnostic accuracy against standard weight-based uroflowmetry in male patients with lower urinary tract symptoms (LUTS), as part of the SAVE (Smartphone-AI Videouroflowmetry Evaluation) Study. Methods: A prospective study was conducted involving 103 male patients aged 48–79 years with LUTS. Each participant underwent standard weight-sensor uroflowmetry and simultaneous video recording of voiding using a smartphone. A proprietary AI algorithm was employed to extract the maximum flow rate (Qmax), while voided volume was visually estimated by an independent expert blinded to the results of standard uroflowmetry. Agreement between methods was assessed using Lin’s concordance correlation coefficient (CCC), Bland–Altman plots, absolute percentage error (APE), and responsiveness indices. Results: AI-VUF showed near-perfect concordance with standard uroflowmetry: CCC was 0.968 for Qmax and 0.992 for voided volume. Mean difference in Qmax was 0.04 mL/s; median APE was 4.0%. Bland–Altman analysis revealed narrow limits of agreement (– 3.66 to + 3.74 mL/s). No significant proportional bias was detected. The system required no calibration and functioned solely on consumer-grade smartphones. Conclusion: This pilot SAVE study demonstrates that AI-assisted videouroflowmetry via smartphone achieves diagnostic performance comparable to standard uroflowmetry. This method offers a promising, scalable solution for remote or office-based assessment of voiding function, with potential to improve access to urological diagnostics. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.