How to write a Kling prompt: a four-line template

A working Kling prompt template: subject, action, camera, constraints. Text-to-video in NEIRON without a VPN or extra account.

How we checked

The same starter still and four prompt lines in Kling inside NEIRON. We watched camera and face after dropping “8k masterpiece” and keeping subject, verb, camera, and constraints.

Table: what to write in Kling

ModelFaceText → videoPhoto → videoPriceSpeed
KlingMid without a stillStrong text-to-videoStarts from a frameStudio versionSocial length
SeedanceBetter face joinWeaker from scratchAnimate a stillMini / 2.0Character
VeoCinema, not a seriesComplex sceneRarely a daily millCostlierOne take

See also

A Kling prompt is not a movie synopsis. The model needs camera, subject, motion, and what must stay still. Below is a working template we use in NEIRON — without a VPN or a separate Kling account.

This page is about writing the prompt. If you already have a still and want to animate it, open the image-to-video guide. If you need a cinematic look from text, compare Kling and Veo after you have a first clip.

The four-line template

  1. Subject: who or what is on screen, age, clothes, one distinctive detail.
  2. Action: one verb — walks, turns, hair moves, steam rises. Not five plots at once.
  3. Camera: slow dolly in, static, handheld, pan left. Kling follows camera words more reliably than “cinematic masterpiece”.
  4. Constraints: keep the face, no morphing, no extra people, natural light.

Example that actually runs:

A woman in a beige coat on a wet city street at dusk. She walks toward camera, coat hem moves. Slow dolly in, 35mm, shallow depth of field. Keep identity, no extra characters, natural reflections on asphalt.

What to drop from the prompt

  • Score lists like “8k, trending on artstation, award winning”. They waste tokens and rarely change motion.
  • A second plot in the same clip: rain plus dance plus a car crash. Split into two generations.
  • Contradictions: “static camera” and “fast orbit” together.

Text-to-video vs image-to-video

From text, Kling invents the first frame. Write light and location or the face will drift. From a photo, describe only what should move — hair, cloth, camera — and lock the rest. For a still you care about, start with image-to-video.

Guide: animate a photo · Kling · Video generator

Where to run it

In NEIRON open Kling, paste the four lines, pick duration in the studio, check the credit cost before you hit generate. Same balance as Nano Banana and Suno — no extra Kling login.

Open the studio · Kling vs Veo · Seedance vs Kling