Objective: To construct an intelligent teaching model for clinical microbiology laboratory testing based on artificial intelligence tools, aiming to enhance students' professional skills and clinical thinking abilities. Methods: The “scenario analysis-requirement refinement-intelligent empowerment-integrated practice”four-step construction method was adopted, using bloodstream infection pathogen identification as a case study. Interactive decision trees were generated by optimizing prompts via Kimi, and the “Blood Culture Identification Assistant” intelligent agent was developed on the Coze platform. A controlled teaching experiment involving 60 participants was conducted. Results: The experimental group demonstrated significant superiority over the control group in identification accuracy, efficiency, rare pathogen recognition capability, and clinical thinking scores (p < 0.01). Conclusion: This model successfully achieves the deep integration of AI tools and specialized teaching, providing a replicable practical solution for the digital reform of medical laboratory education.