Prompt

A prompt is the written instruction you give a generative model that describes the output you want it to produce. It is the steering wheel of any text-to-image or text-to-video system: the model knows how to render almost anything, and the prompt is how you tell it which “anything” you mean. A good prompt is specific and concrete rather than vague. “A barista in a sunlit Brooklyn cafe, warm morning light, handheld camera, vertical” gives the model far more to work with than “a coffee video.” Effective prompts name the subject, setting, lighting, mood, camera behavior, and format, because each detail you leave out is one the model decides for you. Prompting has quietly become a practical skill: the same tool produces forgettable or striking results depending entirely on how clearly you ask. The good news is that it is learnable, and small edits reliably move the output toward what you want.

Anatomy of a strong prompt

Think in layers: subject, scene, lighting, motion, framing. Add what matters, cut what does not, and change one variable at a time so you can tell which word moved the result.

Prompt versus script

A prompt describes the shot; a video script is the words spoken inside it. In a tool with a talking avatar, you often write both: the prompt sets the visual, the script sets the dialogue. Teswir turns a single prompt into a finished cinematic short.