Flash Face is implemented using the Google Gemini 2.0 Flash vision model for locating and recognizing student faces in classroom photos under various settings. The model matches known faces to a live database of names, registration numbers and timestamps, making attendance records. A login module which is held secure so as to ensure that only authorized users have access and keeping the data privacy in place. And where necessary, a manual override can be used. The frontend provides a responsive interface for taking pictures and viewing logs. This was created using Next.js, React, TypeScript, ShadCN UI, and Tailwind CSS. For efficient image processing and result generation, the backend integrates Next.js Server Actions and Genkit workflows. This method reduces human error and modernizes classroom management by offering a scalable, dependable, and real-time attendance management system. The experimental evaluation of this was proved to have an accuracy between 95%–97% and worked more efficiently when compared to the models trained with traditional algorithms.