Although the teaching of narrative texts in primary school is well researched, there is a lack of insight into how visual models and multimodal prompts are used by teachers to convey genre-specific knowledge. Therefore, we conducted a multimodal study of the teaching practices in two first-grade classrooms during joint re-tellings of the folktale Little Red Riding Hood and subsequent interactions around narrative genre. In both cases, the teachers used the graphical model The Story Face as well as a whiteboard canvas in their orchestrations. Data was collected in the form of audio and video recordings. Underpinned by a social semiotic framework combined with Bernstein’s concept framing, the analyses revealed that both teachers focused on story events and the story’s macrostructure while displaying different orientations in the use of verbal language and visual representations. This resulted in different emphases on either story-specific or more general features of narrative genre. Furthermore, the students showed an interest in iconic and suspense-building story dialogue, but this aspect was generally de-emphasized by the teachers’ use of verbal language and visual resources. Based on these findings, we discuss the significance of studying teachers’ differing orchestrations through overlapping modes.