Kling 3 API

Kling 3.0 is Kuaishou's third-generation AI video model, and the Kling 3.0 API brings text to video, image to video, and reference clips together at one endpoint. Generate cinematic AI videos with native audio generation, multi shot storytelling, and subject consistency, then receive every generated video through async jobs and webhooks for fast video generation behind one AI API and key. The Kling 3 API runs on Apiframe with no separate Kling AI account to manage.

Integrate Kling 3.0 with a single API call — one key, one unified endpoint, and shared billing across every model on Apiframe.

Get API Key Use in Apiframe Studio View Kling 3.0 API Docs

model: "kling-3.0"

What's special about Kling 3.0

AI Director and multi shot storytelling

The AI Director, Kling AI's headline feature, plans a sequence before the first frame and creates videos with up to six distinct shots in a single generation, each with its own framing, camera angles, and narrative flow. The Kling 3.0 API keeps spatial continuity automatically, so one prompt returns an edited multi shot sequence rather than a single take, a real leap for multi shot video generation. Multi shot support means ads, trailers, and social clips ship from a single prompt.

Text to video and image to video

Start from text prompts for text to video, or animate static images with image to video. Kling 3.0 also supports reference to video: upload images or a short reference video and the Kling 3.0 API carries that look into the generated video. One integration covers text to video, image to video, and reference to video, so every mode of video generation lives behind a single Kling video model.

Upgraded native audio generation

Kling 3.0 Omni co-generates synchronized sound with the video, so the Kling 3.0 API delivers native audio generation: music, sound effects, and environmental soundscapes in the same pass. It also adds natural lip sync and multilingual lip sync across five languages, with per-character voices for multilingual dialogue. You get native audio without stitching a separate track onto the generated video.

Kling video with cinematic camera movement

Direct the camera, camera angles, and shot transitions with natural language, and the Kling 3.0 API follows the cinematic language you describe. Kling video renders realistic motion and scene transitions while holding visual continuity from shot to shot. The result is dynamic video content that reads like it was edited, raising the bar for cinematic video generation.

High fidelity generated video at native 4K

The Kling 3.0 API renders high fidelity video at native 4K rather than upscaling, with stronger physics for water, fabric, and human anatomy. It is tuned to balance quality and speed, and the Pro tier pushes detail further for complex scenes and high fidelity video generation. Faster processing keeps production ready video moving through busy pipelines.

Subject consistency and consistent characters

The Kling 3.0 API locks a character's appearance and voice from reference images or a short clip, so subject consistency holds across shots and scene changes. Multi character coreference keeps consistent characters distinct when several appear together, giving reliable character consistency for series and campaigns. This visual continuity is what makes multi shot narratives feel coherent.

Accurate text and on screen text

The Kling 3.0 API keeps logos, signage, and captions sharp, so accurate text and on screen text stay legible across shots. It renders text in multiple languages, which helps work that ships globally. Precise control over text descriptions and text prompts means the words you write appear the way you intend.

Start frame and end frame control

Supply a start frame, an end frame, or both, and Kling 3.0 generates the motion between them, frame-accurate video generation with no guesswork. Provide reference images for an image to video start, or set an end frame to guide where the generated video lands. End frame guidance is ideal for loops, product reveals, and clean cuts.

Access Kling 3.0 through one AI API

Apiframe gives you API access to Kling 3.0 behind one AI API and key, with no separate Kling AI account. The Kling 3.0 API uses async jobs: send a request, get a task id, and poll or receive a webhook when the videos generated are ready. The same AI API also runs other Kling video models for video generation and dozens more, so you call it the way you call everything else.

Kling AI built for production workflows

Kling AI ships fast, and the Kling 3.0 API exposes the controls production workflows need: video duration, negative prompts, and prompt adherence through a cfg-scale style control. Generous rate limits and faster processing suit busy pipelines, and you can bring your own prompt or a saved template. Standard and Pro options let you trade cost for fidelity per request.

Creative video production, end to end

From a single prompt to a finished cut, Kling 3.0 covers creative video production without a separate editor. It handles video editing tasks like extending a shot, swapping elements, or adjusting the look, and keeps the result consistent. The Kling 3.0 API turns text descriptions into polished AI videos in one place.

Built for real use cases

Teams reach for Kling video to make AI videos at scale: ads, explainers, product demos, and dialogue scenes. One Kling 3.0 API call can return a multi shot sequence, so it creates videos faster than stitching tools together. Whether you need a quick text to video clip or a multi shot narrative, Kling 3.0 fits creative projects and production pipelines, and pairs well with models like Seedance 2.0.

Made with Kling 3.0

A few outputs generated through the Kling 3.0 API on Apiframe.

Text to video: a 5-second cinematic clip of a barista pouring latte art, warm cafe light, slow push-in.

Image to video: animate this sneaker rotating on a pedestal, studio lighting, a branded logo staying sharp and legible.

Multi shot storytelling: a two-shot dialogue between two friends at a table, lip-sync and ambient audio.

Image to video from a portrait: a gentle head turn and blink in soft window light, with an end frame on a smile.

Multi shot mode: a wide city street cutting to a close-up of a neon sign, continuous mood and clean cuts.

Reference to video: match the realistic motion of a waterfall, flowing water and mist, with sound effects and native audio.

Overview

Endpoint: POST /v2/videos/generate
Model ID: kling-3.0
Params key: klingParams
Modality: Video
Provider: Kuaishou
Avg. completion: ~150s

Capabilities

Aspect ratios1:1, 9:16, 16:9

Durations3s, 5s, 8s, 10s, 15s

Image inputSupported

AudioSupported

Avg. time~150s

Quick start

Send a single POST /v2/videos/generate request with your API key to generate with Kling 3.0. The call returns a jobId you can poll or receive via webhook.

curl -X POST https://api.apiframe.ai/v2/videos/generate \
  -H "X-API-Key: afk_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
        "prompt": "a cinematic sunrise over a futuristic cityscape, smooth camera push-in",
        "model": "kling-3.0",
        "klingParams": {
            "start_image": "https://example.com/input.jpg",
            "mode": "pro",
            "generate_audio": false,
            "end_image": "https://example.com/input.jpg"
        }
    }'

import requests

response = requests.post(
    "https://api.apiframe.ai/v2/videos/generate",
    headers={
        "X-API-Key": "afk_your_api_key_here",
        "Content-Type": "application/json",
    },
    json={
        "prompt": "a cinematic sunrise over a futuristic cityscape, smooth camera push-in",
        "model": "kling-3.0",
        "klingParams": {
            "start_image": "https://example.com/input.jpg",
            "mode": "pro",
            "generate_audio": False,
            "end_image": "https://example.com/input.jpg"
        }
    },
)
print(response.json())  # { "jobId": "...", "status": "QUEUED" }

const response = await fetch("https://api.apiframe.ai/v2/videos/generate", {
  method: "POST",
  headers: {
    "X-API-Key": "afk_your_api_key_here",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "prompt": "a cinematic sunrise over a futuristic cityscape, smooth camera push-in",
    "model": "kling-3.0",
    "klingParams": {
      "start_image": "https://example.com/input.jpg",
      "mode": "pro",
      "generate_audio": false,
      "end_image": "https://example.com/input.jpg"
    }
  }),
});
const { jobId } = await response.json();
console.log(jobId);

Response & job lifecycle

Generation is asynchronous. A successful submission returns 202 Accepted with a jobId. Poll GET /v2/jobs/{id} (or supply a webhook_url) until the status is COMPLETED; the result field then holds the output URL(s).

1. Submission response (202)

{
  "jobId": "b2c3d4e5-f6a7-8901-bcde-f23456789012",
  "status": "QUEUED"
}

2. Poll for the result

curl https://api.apiframe.ai/v2/jobs/JOB_ID \
  -H "X-API-Key: afk_your_api_key_here"

import requests, time

while True:
    job = requests.get(
        "https://api.apiframe.ai/v2/jobs/JOB_ID",
        headers={"X-API-Key": "afk_your_api_key_here"},
    ).json()
    if job["status"] in ("COMPLETED", "FAILED"):
        break
    time.sleep(2)
print(job["result"])

let job;
do {
  await new Promise((r) => setTimeout(r, 2000));
  job = await fetch("https://api.apiframe.ai/v2/jobs/JOB_ID", {
    headers: { "X-API-Key": "afk_your_api_key_here" },
  }).then((r) => r.json());
} while (job.status !== "COMPLETED" && job.status !== "FAILED");
console.log(job.result);

Input schema

Request parameters accepted by the Kling 3.0 endpoint. Model-specific options are nested under the params object shown below.

Parameter	Type	Required	Default	Allowed / range	Description
prompt	string	required	—	—	Text description of what to generate.
model	string	required	"kling-3.0"	"kling-3.0"	The model identifier for this endpoint.
klingParams.start_image	string (URL)	optional	—	—	Use a still as the first frame of the clip.
klingParams.mode	string	optional	"pro"	"standard", "pro", "4k"	Mode
klingParams.generate_audio	boolean	optional	false	—	Generate audio
klingParams.end_image	string (URL)	optional	—	—	Optional last frame to control the ending.
klingParams.negative_prompt	string	optional	—	—	Negative prompt
klingParams.multi_prompt	string	optional	—	—	Multi-prompt

Frequently Asked Questions

Common questions about the Kling 3.0 API.

What is Kling 3.0?

Kling 3.0 is Kuaishou's third-generation AI video model, released in February 2026, built on a unified multimodal architecture for video generation. The Kling Video 3.0 series spans Standard and Pro variants plus a Kling 3.0 Omni model for native audio. It creates videos from text to video, image to video, and reference clips in one model, and it is the flagship release from Kling AI.

What is the AI Director feature in Kling 3.0?

The AI Director lets a single prompt generate up to six shots in a single generation, each with its own duration, framing, and camera moves. It plans narrative flow and handles continuity automatically, so the Kling 3.0 API returns a multi shot sequence instead of one take. This multi shot mode is what gives Kling 3.0 its multi shot narratives.

How do I call the Kling 3.0 API?

Get an api key, store it in an environment variable, and send a POST request to the Kling 3.0 API. Apiframe gives you Kling video behind one AI API, so you call this Kling AI model the way you call every other video generation model. The request returns a task id; poll it or use a webhook to fetch the generated video.

Does Kling 3.0 generate audio?

Yes. Kling 3.0 Omni does native audio generation, so the Kling 3.0 API ships every generated video with synchronized native audio: music, sound effects, and ambience. It also adds natural lip sync and multilingual lip sync across five languages for multilingual dialogue.

What is the difference between text to video and image to video?

Text to video starts from text prompts only, while image to video animates a still image you provide. For image to video you can provide reference images and an end frame, and reference to video carries motion from a short reference video into the new clip.

How does multi shot storytelling work?

This mode lets the Kling 3.0 API generate a multi shot sequence in one clip, with shot transitions handled for you. The AI Director keeps continuity, so this multi shot mode reads as a single coherent story. Kling 3.0 offers multi shot support and supports multi shot storytelling up to six shots per clip.

What video duration and aspect ratio are supported?

Video duration runs from 3 to 15 seconds with flexible control, at native resolution up to 4K. Set the aspect ratio and duration per request to balance quality and file size for production pipelines.

Can Kling 3.0 keep characters consistent?

Yes. Kling 3.0 locks a character's appearance and voice from reference images or a few frames, giving subject consistency and character consistency across shots. Multi character coreference keeps each character distinct when several share a scene.

How do I control the camera and the look?

Describe camera movement, camera angles, and cinematic language in natural language, and Kling 3.0 follows it closely with strong prompt adherence. You keep control over visual style and realistic motion for dynamic video content.

Can Kling 3.0 edit or extend a video?

Yes. Kling 3.0 handles video editing inside the same model: extend a shot, replace elements, or restyle a clip without a separate editor. The generated video keeps continuity, which suits demanding production pipelines.

What are the rate limits and how fast is Kling video generation?

Apiframe applies generous rate limits, and faster processing on Kling 3.0 keeps videos generated quickly even in high volume pipelines. Async jobs and webhooks mean your server is not blocked while the video renders.

Which Kling Video 3.0 tiers and models are available?

The Kling Video 3.0 series offers Standard and Pro tiers, plus Kling 3.0 Omni for native audio and voice cloning. Standard balances cost and speed; Pro raises fidelity for complex scenes and high fidelity video.

How do I get an api key and free credits?

Create an Apiframe account to get an api key and free credits, then call Kling 3.0 right away. New users can test text to video and image to video before adding billing.

How do I track a job and its task id?

Each call to the Kling 3.0 API returns a task id. Poll the task endpoint, or set a webhook, and the video url for the generated video arrives when the job completes.

Where can you access Kling 3.0?

You can access Kling 3.0 through the Apiframe API, as well as the Kling AI app. Apiframe gives you one AI API and key for Kling video alongside other models, so creative projects stay on a single bill.

Still have questions?

Contact Support Join Discord

Start building with the Kling 3.0 API

Get your API key and integrate Kling 3.0 in minutes — Pay-as-you-go.

Free credits to start

One API for every model

Webhooks, SDKs & idempotency

No provider account required

Get Started Free Read Documentation

Questions? Join our Discord or contact sales.