Subject
State who or what the scene is about before you add style words or camera notes.
LTX 2.3 Prompt Template
If you already know the kind of clip you want but do not want to start from a blank prompt box every time, a reusable prompt template is the practical middle ground.
This page gives you a ready structure for LTX 2.3 prompts, explains what each part should do, and shows how to adapt one template across cinematic, product, and short-form scenes without losing clarity.
A useful prompt template is not a long block of fixed text. It is a reusable structure with enough guidance to keep your scene coherent while leaving room to swap in a new subject, movement, setting, or style.
State who or what the scene is about before you add style words or camera notes.
Name the motion or event clearly so the clip has a reason to move.
Add one camera instruction such as slow push-in, side tracking, or fixed frame.
Finish with visual tone, lighting, or commercial feel only after the core scene is clear.
[Main subject] in [setting], [clear action or motion], [camera behavior], [lighting or atmosphere], [visual tone or style], [one detail the viewer should notice].
[Product] on [surface or scene], [product motion or reveal], [camera move], [lighting setup], premium commercial mood, clean background, focus on [hero detail].
[Character] in [location], [gesture or action], [camera distance and movement], [lighting feel], cinematic tone, emphasis on [expression or body language].
[Object or person] doing [single satisfying action], quick [camera angle or move], crisp lighting, clean background, short-form ad style, attention on [visual payoff].
If you want fully written examples instead of reusable blanks, the best companion pages are prompt examples and best prompts.
Do not rewrite the whole prompt. Change the subject, setting, and hero detail first while keeping the same order.
If the output is close, only test one new action or one new camera move so you can see what changed the result.
Once a template variant works, save it as a repeatable starting point in aicovea instead of rebuilding the prompt every session.
Related Guides
No, as long as the template preserves clear scene logic. The template gives structure; the specificity comes from the subject, action, and visual details you plug into it.
The biggest mistake is treating every field like a style bucket. Start with the main subject and motion first, then add camera and look.
A browser workflow is usually the fastest starting point because you can iterate on the structure quickly before you commit to a local or developer-heavy setup.