model-comparison

name	model-comparison
description	Build a controlled academy comparison of animation or generation models using the same task inputs and evaluation dimensions. Use when teaching tradeoffs between supported multimodal workflows. Do not use when only one model is being documented or when current platform facts cannot be verified.

Model comparison

• Define one common task and reuse the same source inputs across models where possible.
• Record model version date duration aspect ratio and material settings.
• Compare dimensions relevant to the lesson such as continuity motion prompt adherence runtime and cost.
• Separate measured observations from subjective preference.
• Note platform limitations and verify current fast-changing facts before publishing them.
• Avoid declaring a universal winner from one test case.
• Run lint and build after adding comparison content.