The development of modern science and technology and multimodal discourse analysis theory has led to the transformation of classroom teaching to multimodal teaching mode. Language, image, motion and other symbolic resources convey information through multiple sensory channels such as auditory, visual and tactile, which promote students’ reception of information resources and improve learning effect. This study selects a demonstration video of the elementary-level comprehensive course of international Chinese teaching, uses the software ELAN5.4 to annotate and calculate the modal symbols used in the class, analyzes the application of modalities in the process of teaching and explores how modalities work together in teaching. This study also divides the video into many images to explore the multimodal combination. The research results show that the PPT is the most commonly used modal symbol and the auditory mode is the main mode; the main modal symbols and main modes used in different teaching parts are diverse; this video is divided into 82 images based on meaning and a total of 39 multimodal combinations are used; the most frequently used multimodal combination is teacher’s discourse, students’ discourse, PPT (image and text) and sign language; there are eight commonly used multimodal combinations.