live
MODEL IDklingai:kling-image@4.0

Kling IMAGE 4.0

Kling AI
by

Kling IMAGE 4.0 is an image model for text-to-image and reference-guided generation at native 2K, 4K, and 8K, with natural lighting and skin and material detail close to a camera photo. It renders up to 2,000 characters of legible text in seven languages, combines up to 10 reference images and elements, and plans series of up to 10 images that keep subjects and style consistent. It also designs e-commerce product pages and websites as a single tall image, which suits storyboards, slide decks, product shoots, and campaign visuals.

Kling IMAGE 4.0

Image series

How to generate a planned set of 2 to 10 consistent images with Kling IMAGE 4.0: storyboards, slide decks, product image sets and lookbooks from one request.

Introduction

A storyboard or a slide deck is a set of images that have to match. Generated one at a time, each frame drifts: the courier's jacket changes shade, the slide titles move. With settings.seriesLength, Kling IMAGE 4.0 plans the whole set in one request and keeps the subject and the style consistent across it. The six frames below came back from a single call.

The yellow jacket and backpack are the continuity anchors. The prompt describes them once and then asks for them "identical in every frame", which is what a storyboard artist would hold steady across the cuts.

This guide covers the request, how to plan the images inside one prompt, slide decks, letting the model pick the count, and building a series from reference images.

The request

A series is an ordinary request with settings.seriesLength set to the number of images. This is the request behind the storyboard:

Try in Playground
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'klingai:kling-image@4.0',
  positivePrompt: 'A six-frame storyboard for a 30-second commercial for a food delivery app, all frames 16:9. Rainy evening in a big North American city, wet streets reflecting neon and shop lights, cool blue shadows with warm amber highlights, cinematic film still look with fine grain. Every sign and screen in the frames reads in English. The courier is a woman in her twenties with a short black bob, in a bright yellow rain jacket with a reflective stripe and a yellow insulated delivery backpack, riding a black electric bicycle. Keep her face, her jacket and her backpack identical in every frame. 01 | Wide establishing shot: she rides out of a narrow alley into a busy avenue, rain streaking through the headlights. 02 | Medium tracking shot from the side: she weaves past a stopped bus, water spraying from her front wheel. 03 | Close-up: her gloved hand checks the order on a phone mounted on the handlebars, the screen showing a delivery map with a route line and glowing on her wet face. 04 | Low angle: she locks the bicycle outside an apartment building and runs up the front steps. 05 | Over-the-shoulder shot: a door opens and a tired man in a hoodie takes the warm paper bag from her with a grateful smile. 06 | Wide final shot: she rides off into the rain, small in the frame, the avenue glowing ahead of her. No logos.',
  width: 2720,
  height: 1536,
  settings: {
    seriesLength: 6
  }
})
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "klingai:kling-image@4.0",
            "positivePrompt": "A six-frame storyboard for a 30-second commercial for a food delivery app, all frames 16:9. Rainy evening in a big North American city, wet streets reflecting neon and shop lights, cool blue shadows with warm amber highlights, cinematic film still look with fine grain. Every sign and screen in the frames reads in English. The courier is a woman in her twenties with a short black bob, in a bright yellow rain jacket with a reflective stripe and a yellow insulated delivery backpack, riding a black electric bicycle. Keep her face, her jacket and her backpack identical in every frame. 01 | Wide establishing shot: she rides out of a narrow alley into a busy avenue, rain streaking through the headlights. 02 | Medium tracking shot from the side: she weaves past a stopped bus, water spraying from her front wheel. 03 | Close-up: her gloved hand checks the order on a phone mounted on the handlebars, the screen showing a delivery map with a route line and glowing on her wet face. 04 | Low angle: she locks the bicycle outside an apartment building and runs up the front steps. 05 | Over-the-shoulder shot: a door opens and a tired man in a hoodie takes the warm paper bag from her with a grateful smile. 06 | Wide final shot: she rides off into the rain, small in the frame, the avenue glowing ahead of her. No logos.",
            "width": 2720,
            "height": 1536,
            "settings": {
                "seriesLength": 6
            }
        })


asyncio.run(main())
curl https://api.runware.ai/v1 \
  -H "Authorization: Bearer $RUNWARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '[
    {
      "taskType": "imageInference",
      "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
      "model": "klingai:kling-image@4.0",
      "positivePrompt": "A six-frame storyboard for a 30-second commercial for a food delivery app, all frames 16:9. Rainy evening in a big North American city, wet streets reflecting neon and shop lights, cool blue shadows with warm amber highlights, cinematic film still look with fine grain. Every sign and screen in the frames reads in English. The courier is a woman in her twenties with a short black bob, in a bright yellow rain jacket with a reflective stripe and a yellow insulated delivery backpack, riding a black electric bicycle. Keep her face, her jacket and her backpack identical in every frame. 01 | Wide establishing shot: she rides out of a narrow alley into a busy avenue, rain streaking through the headlights. 02 | Medium tracking shot from the side: she weaves past a stopped bus, water spraying from her front wheel. 03 | Close-up: her gloved hand checks the order on a phone mounted on the handlebars, the screen showing a delivery map with a route line and glowing on her wet face. 04 | Low angle: she locks the bicycle outside an apartment building and runs up the front steps. 05 | Over-the-shoulder shot: a door opens and a tired man in a hoodie takes the warm paper bag from her with a grateful smile. 06 | Wide final shot: she rides off into the rain, small in the frame, the avenue glowing ahead of her. No logos.",
      "width": 2720,
      "height": 1536,
      "settings": {
        "seriesLength": 6
      }
    }
  ]'
runware run klingai:kling-image@4.0 \
  positivePrompt="A six-frame storyboard for a 30-second commercial for a food delivery app, all frames 16:9. Rainy evening in a big North American city, wet streets reflecting neon and shop lights, cool blue shadows with warm amber highlights, cinematic film still look with fine grain. Every sign and screen in the frames reads in English. The courier is a woman in her twenties with a short black bob, in a bright yellow rain jacket with a reflective stripe and a yellow insulated delivery backpack, riding a black electric bicycle. Keep her face, her jacket and her backpack identical in every frame. 01 | Wide establishing shot: she rides out of a narrow alley into a busy avenue, rain streaking through the headlights. 02 | Medium tracking shot from the side: she weaves past a stopped bus, water spraying from her front wheel. 03 | Close-up: her gloved hand checks the order on a phone mounted on the handlebars, the screen showing a delivery map with a route line and glowing on her wet face. 04 | Low angle: she locks the bicycle outside an apartment building and runs up the front steps. 05 | Over-the-shoulder shot: a door opens and a tired man in a hoodie takes the warm paper bag from her with a grateful smile. 06 | Wide final shot: she rides off into the rain, small in the frame, the avenue glowing ahead of her. No logos." \
  width=2720 \
  height=1536 \
  settings.seriesLength=6
{
  "taskType": "imageInference",
  "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
  "model": "klingai:kling-image@4.0",
  "positivePrompt": "A six-frame storyboard for a 30-second commercial for a food delivery app, all frames 16:9. Rainy evening in a big North American city, wet streets reflecting neon and shop lights, cool blue shadows with warm amber highlights, cinematic film still look with fine grain. Every sign and screen in the frames reads in English. The courier is a woman in her twenties with a short black bob, in a bright yellow rain jacket with a reflective stripe and a yellow insulated delivery backpack, riding a black electric bicycle. Keep her face, her jacket and her backpack identical in every frame. 01 | Wide establishing shot: she rides out of a narrow alley into a busy avenue, rain streaking through the headlights. 02 | Medium tracking shot from the side: she weaves past a stopped bus, water spraying from her front wheel. 03 | Close-up: her gloved hand checks the order on a phone mounted on the handlebars, the screen showing a delivery map with a route line and glowing on her wet face. 04 | Low angle: she locks the bicycle outside an apartment building and runs up the front steps. 05 | Over-the-shoulder shot: a door opens and a tired man in a hoodie takes the warm paper bag from her with a grateful smile. 06 | Wide final shot: she rides off into the rain, small in the frame, the avenue glowing ahead of her. No logos.",
  "width": 2720,
  "height": 1536,
  "settings": {
    "seriesLength": 6
  }
}
Response
[
  {
    "taskType": "imageInference",
    "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
    "imageUUID": "1b7e3c9a-4d28-4f61-8e05-9a3c7d2f6b14",
    "imageURL": "https://im.runware.ai/image/os/a14d18/ws/2/ii/1b7e3c9a-4d28-4f61-8e05-9a3c7d2f6b14.jpg"
  },
  {
    "taskType": "imageInference",
    "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
    "imageUUID": "6c2a8f4d-1e93-4b57-a0d6-3f8b1e5c9a72",
    "imageURL": "https://im.runware.ai/image/os/a14d18/ws/2/ii/6c2a8f4d-1e93-4b57-a0d6-3f8b1e5c9a72.jpg"
  },
  {
    "taskType": "imageInference",
    "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
    "imageUUID": "3e9d5b1f-7a46-4c82-b1e9-5d4a2c8f7e30",
    "imageURL": "https://im.runware.ai/image/os/a14d18/ws/2/ii/3e9d5b1f-7a46-4c82-b1e9-5d4a2c8f7e30.jpg"
  },
  {
    "taskType": "imageInference",
    "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
    "imageUUID": "9f1c6e3a-2b85-4d07-a4c3-8e6b2d9f1a58",
    "imageURL": "https://im.runware.ai/image/os/a14d18/ws/2/ii/9f1c6e3a-2b85-4d07-a4c3-8e6b2d9f1a58.jpg"
  },
  {
    "taskType": "imageInference",
    "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
    "imageUUID": "4a8b2e7c-5f19-4e36-9d02-7c1f4b8e3a65",
    "imageURL": "https://im.runware.ai/image/os/a14d18/ws/2/ii/4a8b2e7c-5f19-4e36-9d02-7c1f4b8e3a65.jpg"
  },
  {
    "taskType": "imageInference",
    "taskUUID": "8d2f4a61-7c3e-4b19-a5d0-2e9c6f1b8a43",
    "imageUUID": "7d3f9a5b-8c21-4a74-b6e8-1f5c3a9d7b26",
    "imageURL": "https://im.runware.ai/image/os/a14d18/ws/2/ii/7d3f9a5b-8c21-4a74-b6e8-1f5c3a9d7b26.jpg"
  }
]
  • settings.seriesLength takes a count from 2 to 10, or "auto" to let the model choose one.
  • The images come back in series order, one entry each, all at the size the request set.
  • A series is planned as one set, so it always comes from a single request with numberResults at 1.

Planning each image

The storyboard prompt has two parts: a brief for the whole set and one line per frame. The brief carries what every frame shares, from the courier to the light. Each numbered line then carries only what is new in that frame, its shot size and its action.

Number the lines to match the series. "01 |" through "06 |" gives each frame its own instruction and fixes the order the images come back in. Without numbered lines the model plans the set itself from the brief, which suits a product set. A story needs its beats in a fixed order, so it gets the lines.

Ask for consistency in words. "Keep her face, her jacket and her backpack identical in every frame" names the parts that must not drift. The model plans the set as a whole, and the sentence tells it which details to hold fixed across it.

Name the language of any text in the scene. Without "Every sign and screen in the frames reads in English", the same storyboard came back with Chinese neon signs and a Chinese phone screen, even with the city set in North America.

Slide decks

Each slide of a deck is one image of the series, so a five-slide pitch is a request with seriesLength 5 at 16:9. Kling recommends building the prompt in five parts:

[Topic and audience] + [Number of slides and aspect ratio] + [Visual style]
+ [Optional: slide-by-slide outline] + [Consistency requirements]

The consistency line is what makes it a deck. "Keep the color palette, the title style, the slide numbers in the bottom right and the small wordmark in the top left identical on every slide" lists the elements a template would repeat, so the five images read as one file. The outline is optional: give only the topic and the style, and the model plans the slides itself.

Say what every screen shows. A phone or laptop mockup left undescribed gets interface text the model writes itself, whatever language the prompt asks for. The cover describes its phone as a calendar grid of colored shift blocks with no words, and the screen came back without any.

Letting the model choose the count

With seriesLength set to "auto", the model decides how many images the brief needs, between 2 and 10. That suits a product listing, where the right set depends on the product rather than on a fixed slot count. The bottle below went in as one packshot:

Describe the goal of the set rather than each image. "Choosing the shots that best show what a buyer needs to see" hands the planning to the model, and the list of shot types after it marks what the set should cover. Read the number of images from the response, since it changes with the brief.

Series from reference images

A series takes inputs.referenceImages like any request, so one model and one garment become a whole lookbook. The man and the coat below went in as two references, shown once above the four images they produced:

Say what each reference contributes and ask for it "in every image". The prompt keeps the face, hair and build from the first reference and the cut, color, buttons and lapels from the second, and the numbered lines only change the pose and the place. How to assign roles to references in general is covered in reference images.

Tips

  1. Split the prompt into a brief and numbered lines. The brief holds for every image, each line adds only what changes.

  2. Name the details that must not drift. "Identical in every frame" next to the jacket or the slide template keeps them fixed.

  3. Match the line count to seriesLength. Six numbered shots for six images keeps each instruction tied to its frame.

  4. Use "auto" when the content decides the count. Read the number of images from the response.

  5. Set the size once for the whole series. Every image comes back at the size the request asked for, so a deck at 16:9 stays 16:9 throughout.

  6. Pair a series with references for product and fashion sets. The references carry the product or the person, the numbered lines carry the shots.