Session 1 · One photo, one motion
Students upload a single portrait and generate three different motion styles from it (Module 04). Compare results as a class before anyone writes a prompt from scratch.
Session 2 · Script to scene
Write a five-shot script using the formula in Module 12, generate each shot with Text → Video (Module 05), and assemble them in order.
Session 3 · 60-second faceless explainer
Using avatars and lip sync (Module 06), students script and generate a short explainer on a topic of their choice, with subtitles.
Session 4 · Cross-model critique
Run one identical prompt through Real Motion 3.0 (Module 10) and two guest models (Module 11), then submit a written comparison of realism, motion quality and prompt adherence.
Session 5 · Skill teardown
Pick one Agent skill (Module 09), run it, then reverse-engineer what prompt structure and reference assets it is likely packaging. One paragraph, compared as a class.
Capstone · 30-second branded piece
Combine at least three modules — for example image generation, Act-Motion and voice — into one finished exportable clip with a clear brief and a one-paragraph rationale.
Callout
Grading angle: because generation is fast, weight assessment toward craft decisions — prompt construction, shot choice, iteration discipline — not raw polish, which tracks credit tier more than skill.
That's the last module — time to submit a clip.
See the showcase