Kitchen
The room where a reaction has a reason. Tasting, prepping, burning something — the kitchen hands you a physical action to react to, so the hook doesn't have to carry the whole clip on its own.
Cafe
A table, a cup and a reason to be sitting still. The cafe is the setting that makes a long talking-head clip look like a moment rather than a piece to camera.
Morning bedroom
The highest-intimacy room in the set. Soft light, no makeup, a product opened in bed — this is where a hook stops feeling like an ad and starts feeling like a text from a friend.
Selfie vlog
No location, no props, no excuse — just a face and a first sentence. The hardest format in this set and the one that travels furthest when it lands.
Outdoor street
Motion the viewer can feel. A moving background does a job no static room can — it makes a clip feel like it is going somewhere before you have said what.
Office desk
The only setting in this set built around a second screen. Over-the-shoulder framing turns a laptop into a reason to look — and the thing you are reacting to lives inside it.
Gym
The one setting where the state of your body is the hook. Out of breath, mid-wipe, still shaking — physical aftermath is evidence the viewer can see before you claim anything.
Cozy night
Low light, low energy, low stakes. The setting where a quiet line lands harder than a loud one — and the only one in this set where the register is the whole point.
How this set is organised
Every prompt here is one scene description plus, for 52 of them, the line the creator says on camera. They are sorted by filming setting rather than by emotion or shot type, because the setting is the only one of the three that changes what you have to physically arrange before you press record. A shocked reaction and a smug reaction are the same setup; a shocked reaction in a gym and a shocked reaction in bed are not.
The emotion and shot-format mix is shown on each setting page as a breakdown of what that room actually produced. Those are facets on the page, not separate pages — one prompt should live at one address, and splitting the same 80 clips across three taxonomies would put every one of them at three.
Every clip on these pages is a real render, not a mockup. Where a prompt carries a suggested hook line, that line was written alongside the scene. Where it does not, it came from the premade reaction library inside the app, which stores expressions rather than scripts.
The two lines every prompt shares
The scene changes per prompt. These two do not, so they are stated once rather than repeated eighty times — append both when you copy a scene.
Base style line
Vertical phone UGC, authentic skin texture, natural light, slight handheld motion.
Clean plate line
Clean plate only: no on-screen text, no captions, no watermarks, no logos, no UI overlays.
Each setting page adds one more exclusion specific to that room — the thing that reliably ends up in frame there and should not.
Build these hooks in Versely
The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync in a single app.