<p>Information retrieval tools and metacognitive prompts, as forms of external regulation, provide timely interactivity that can support scientific concept understanding in simulated video lecture environments. This study examined the interactive effects of search types (traditional search vs. generative artificial intelligence (GenAI) search) and metacognitive prompts (structuring prompts vs. problematizing prompts) on students’ scientific concept learning using eye-tracking and screen recording methods. A total of 95 Chinese undergraduate and graduate students (<i>M</i>_age = 20.83, <i>SD</i>_age = 2.23) participated in an online learning task in the domain of gene expression biology. Analysis of variance (ANOVA) showed that participants exhibited higher fixation counts and longer dwell times on the metacognitive prompts area. Under the condition of traditional search combined with problematizing prompts, learners also made more saccades, demonstrated deeper self-monitoring behaviors, and achieved somewhat higher transfer. Together, these findings highlight the benefits of combining traditional search with problematizing prompts to support scientific concept understanding in simulation-based video lectures. Practically, educators should systematically embed well-designed metacognitive prompts, particularly problematizing prompts, into inquiry tasks to guide students in monitoring and reflecting on their learning processes. Furthermore, teachers should support students in effectively using traditional search and GenAI search tools to locate relevant and reliable information that meets their individual needs.</p>