Tiny Lighthouse Keepers

Miniature World Text only · 9:16 · 10s video
Tiny Lighthouse Keepers — 3-frame storyboard, Miniature World AI video prompt example

On a stormy night the great lighthouse lamp is dark — and a crew of tiny keepers in oilskins climbs up to relight it before a little boat out in the black waves loses its way. Copy the full master prompt below and paste it into Google Flow, Veo, Kling or any AI video generator. It includes the start frame, timed video direction, sound and a negative prompt — everything needed to recreate the format.

The master prompt

START FRAME (image): A giant lighthouse lamp room at night seen from keeper-height: the huge dark lens in the centre, three tiny keepers in yellow oilskins climbing the last steps carrying a tiny oil can and cloths, storm clouds through the glass, cinematic.
VIDEO (9:16, 10 seconds): 0–4s — the keepers polish the lens in a circle and one primes the lamp; the wind howls outside. 4–7s — the lamp flickers, sputters, then catches: a great beam sweeps out over the black waves. 7–10s — far below, a tiny boat steadies and turns toward the harbour light; the keepers wave from the gallery rail.
SOUND: howling wind, creaking metal, one deep distant foghorn as the beam lights.
NEGATIVE: keepers falling, horror storm tone, lightning strikes on the tower, text, logos, watermarks.
Hook: A dark lens in a storm in frame 1 — the relight is the promise.

How to use this prompt

1. Copy the promptUse the Copy Prompt button above — the full text includes the start frame, timing, sound and negative prompt.
2. Paste it into your video toolWorks with Google Flow (Veo), Kling, Runway, Luma and similar AI video generators — no photo or upload needed, it is text-only.
3. Generate in 9:16Set a vertical 9:16 format and the 10s duration named in the prompt, then post at your audience's peak time.

Frequently asked questions

Do I need a photo to use the Tiny Lighthouse Keepers prompt?

No. This is a text-only prompt — copy it into your AI video tool exactly as written. If you want a locked-in opening frame, generate a still from the start-frame description first, then run the same text as an image-to-video clip.

Which AI video tools does this prompt work with?

Any modern AI video generator — Google Flow (Veo), Kling, Runway or Luma. Paste the full prompt, set a vertical 9:16 frame and a 10-second duration, and generate. The timed beats, sound notes and negative prompt are written to work as-is in all of them.

How do I keep the miniature scale believable?

The prompt shoots from keeper-height and makes the lens huge in frame, which is what sells the scale. Keep that framing if you adapt it: the moment the camera pulls back to a normal wide shot, the keepers stop reading as tiny and the scene loses its stakes.

Related prompts

← Back to the full prompts library