This study focuses on the practical problems in college English speaking instruction

namely students' difficulty in speaking

unstable expression

and the challenge of balancing fluency and accuracy

and aims to investigate the impact of the multimodal audio-visual-speaking teaching model on the oral English output ability of non-English major university students. In this paper

the "multimodal audio-visual-speaking teaching model" is defined as a teaching model that takes dynamic video as its core medium

integrates visual

auditory

and textual semiotic resources

and promotes oral production through contextualized tasks. The study intends to adopt a quasi-experimental design

selecting two parallel classes from a university as research subjects

with one experimental group and one control group

and implementing a 16-week teaching intervention. Data sources will include pre-test and post-test oral samples

questionnaire surveys

and semi-structured interviews. Corpus analysis and statistical analysis will be combined to examine oral fluency and accuracy. The study mainly addresses three questions: whether multimodal teaching can significantly improve university students' oral English output ability; whether its effects on fluency and accuracy differ and whether such effects accord with the trade-off hypothesis; and what mechanisms underlie its role in promoting oral output. Based on previous studies and the assumptions of this research

the expected conclusion is that

compared with traditional instruction

the multimodal audio-visual-speaking teaching model is more conducive to improving students' oral output level

and that its facilitative effect on fluency may be more direct while also exerting a positive effect on accuracy. Video contexts

nonverbal cues

and textual scaffolding in multimodal environments are expected to reduce learners' cognitive load and optimize the allocation of attentional resources

thereby alleviating the trade-off relationship between fluency and accuracy to a certain extent. The findings may provide a theoretical basis and practical reference for the reform of college English audio-visual-speaking courses and the design of oral teaching activities.