Back to course
References and start frames
How to hand a model a real image instead of describing everything in words.
A prompt can only describe so much — sometimes the fastest way to get an exact look, pose, or product is to show the model a real image instead of describing it. Most models in Dayanica let you attach one or more reference images alongside your prompt (look for the thumbnail row under the prompt box, or the model's own "Attach" slot).
A few video models go further with dedicated "start frame" and "end frame" slots: attach the exact image you want the clip to open on (and, for some models, the image it should end on), and the model animates the motion in between. This is the most reliable way to control exactly how a video begins — far more precise than describing the first frame in words.
Once you've attached a reference, you can @mention it by name right inside your prompt (e.g. "@image1 walking through a neon-lit street") — this is the same mechanism used to bring in a saved Element, covered in the next lesson.
Tip:Not every model accepts references the same way — some only take one image, others take several, and a few don't accept any at all. Dayanica greys out or explains what's unsupported for your current model right in the compose area, so you'll never be guessing.
Log in to track your progress