Imagen

Kling 3.0 API AI Video Generation for Developers

Kling 3.0 API for text-to-video, image-to-video and motion control on one ModelsLab key. POST /api/v7/video-fusion/text-to-video with model_id kling-v3-t2v; billed per second of video from $0.084.

Get Kling 3.0 API Access API Documentation

What is the Kling AI API?

Kling AI, exposed as a developer video API

Kling AI is Kuaishou's family of video generation models. ModelsLab exposes Kling 3.0, Kling V2.5 Turbo, Kling V2.1 and V2 Master and Kling motion control as model_ids on its video endpoints, so you call them with a ModelsLab API key instead of a Kuaishou account.

Text-to-video, image-to-video and motion control are separate endpoints under /api/v7/video-fusion/. Renders are asynchronous: the response carries an id, and you POST /api/v7/video-fusion/fetch/{id} or pass a webhook URL.

When to use the Kling AI API

Kling 3.0 suits short-form social, ads and product video where motion quality matters. For drafts and bulk renders, the open-weight models ModelsLab self-hosts (Wan 2.2, LTX 2.3) are covered by your plan and cost less per clip.

Text-to-video: kling-v3-t2v, kling-v2-5-turbo-t2v
Image-to-video: kling-v3-i2v, Kling-V2-5-Turbo-i2v
Motion control: kling-v3-motion-control, kling-motion-control
Per-second billing from plan usage, then wallet balance; mode and audio scale the rate
Webhook or fetch/{id} for the finished MP4

Trusted by

500M+

API Requests Processed

400K+

Registered Users

5K+

Discord Community Members

300+

Available AI APIs

500

GPUs in Our Own Datacenter

Kling AI API quick start

Every Kling model is a model_id on /api/v7/video-fusion/. Send your API key in the key header or body.

Python — Kling 3.0 text to video

Python

1import requests
2
3response = requests.post(
4    "https://modelslab.com/api/v7/video-fusion/text-to-video",
5    headers={"key": "YOUR_MODELSLAB_API_KEY"},
6    json={
7        "model_id": "kling-v3-t2v",
8        "prompt": "a slow dolly-in on a vintage red sports car at sunset, cinematic 35mm",
9        "duration": 5,
10        "aspect_ratio": "16:9",
11        "mode": "std",
12        "generate_audio": False,
13    },
14)
15
16# Returns an id — poll fetch/{id} or pass a webhook
17print(response.json())

JavaScript — Kling 3.0 image to video

JavaScript

1const response = await fetch(
2  'https://modelslab.com/api/v7/video-fusion/image-to-video',
3  {
4    method: 'POST',
5    headers: {
6      'Content-Type': 'application/json',
7      key: 'YOUR_MODELSLAB_API_KEY',
8    },
9    body: JSON.stringify({
10      model_id: 'kling-v3-i2v',
11      prompt: 'slow orbit around the product, soft studio lighting',
12      init_image: ['https://example.com/product.png'],
13      duration: 5,
14      webhook: 'https://your-app.com/webhooks/kling',
15    }),
16  },
17);
18
19console.log(await response.json());

cURL — Kling 3.0 motion control

bash

1curl -X POST 'https://modelslab.com/api/v7/video-fusion/motion-control' \
2  -H 'Content-Type: application/json' \
3  -H 'key: YOUR_MODELSLAB_API_KEY' \
4  -d '{
5    "model_id": "kling-v3-motion-control",
6    "init_image": "https://example.com/character.png",
7    "init_video": "https://example.com/dance-reference.mp4",
8    "character_orientation": "video",
9    "mode": "std"
10  }'
11
12# Then: POST /api/v7/video-fusion/fetch/{id} with the same key

Kling AI API endpoints

Four endpoints under /api/v7/video-fusion/, one API key.

Text-to-video endpoint

POST /api/v7/video-fusion/text-to-video with model_id (kling-v3-t2v), prompt, duration, aspect_ratio, mode (std, pro or 4k) and generate_audio. Returns an id for the render.

Image-to-video endpoint

POST /api/v7/video-fusion/image-to-video with model_id (kling-v3-i2v), prompt and init_image as an array of image URLs. Kling animates the still while keeping the subject.

Motion control endpoint

POST /api/v7/video-fusion/motion-control with init_image (the character), init_video (the movement to copy), character_orientation (image or video) and optional mode and keep_original_sound.

Fetch and webhooks

POST /api/v7/video-fusion/fetch/{id} until output holds the MP4 URL, or pass webhook in the request and ModelsLab posts the result to you.

Kling runs on the same endpoints, key and fetch flow as Wan, Seedance, Veo and MiniMax, billed per second of output.

Get Started

Kling models and prices on ModelsLab

Kling is billed per second of output: from the usage included in Basic ($21/month) or Standard ($47/month) first, then from wallet balance, and from wallet balance on Open Source Unlimited. The listed rate is standard mode without audio.

model_id	Model	Endpoint	Listed rate
kling-v3-t2v	Kling 3.0	text-to-video	$0.084 per second
kling-v3-i2v	Kling 3.0	image-to-video	$0.084 per second
kling-v3-motion-control	Kling 3.0 Motion Control	motion-control	$0.126 per second
kling-v2-5-turbo-t2v	Kling V2.5 Turbo	text-to-video	$0.042 per second
Kling-V2-5-Turbo-i2v	Kling V2.5 Turbo	image-to-video	$0.042 per second
kling-motion-control	Kling Motion Control	motion-control	from $0.07, scaled by options
kling-v2-1-master-t2v	Kling V2.1 Master	text-to-video	$0.336 per second

Rates from each model's dreambooth_models price rule, September 23 2026. Kling 3.0: mode pro ×1.34, mode 4k ×5.04, generate_audio ×1.5; duration defaults to 3 seconds.

Motion control parameters

Motion control copies the movement in a reference video onto the character in your image. POST /api/v7/video-fusion/motion-control.

Parameter	Values	Effect
init_image	image URL (required)	The character or subject to animate
init_video	video URL (required)	The reference clip whose movement is copied
character_orientation	image or video (required)	Keep the pose framing of the image or follow the video
mode	std or pro	Quality tier; pro costs more per second
keep_original_sound	yes or no	Keep the audio of the reference video
prompt	text, up to 2,500 characters	Optional scene description

How to use the Kling AI API

Integrate AI video generation in three simple steps.

STEP 01

Step 1: Get your API key and fund a wallet

Subscribe to a video plan (from $21/month) and create an API key. Kling is a third-party model, billed per second from the usage included in Basic or Standard first, then from wallet balance; Open Source Unlimited does not cover it.

STEP 02

Step 2: Send the request

POST to /api/v7/video-fusion/text-to-video, image-to-video or motion-control with a Kling model_id, your prompt or reference media, and the key header.

STEP 03

Step 3: Fetch the MP4

The response carries an id. POST /api/v7/video-fusion/fetch/{id} until output holds the video URL, or receive it on your webhook.

Get Kling 3.0 API Key

Why teams use the ModelsLab Kling AI API

Key advantages that set us apart

Kling 3.0, V2.5 Turbo, V2.1 Master and motion control on one key

No Kuaishou account or separate billing relationship

Same endpoints and fetch flow as Wan, Seedance, Veo and MiniMax

Per-second billing with the rate shown per model_id

Webhook callbacks for async delivery

Draft on self-hosted Wan or LTX covered by your plan, render finals on Kling

Our Popular Use Cases

B2B applications powered by the Kling AI API:

Offer text-to-video, image animation and motion control to your users without a Kuaishou integration.

Related developer APIs

Kling AI Playground

Try Kling AI in the online video playground

Text to Video API

Wan, Seedance, Veo and more video models

Image Generation API

Generate reference images for image-to-video flows

Your Data is Secure: GDPR Compliant AI Services

ModelsLab GDPR Compliance Certification Badge

GDPR Compliant

Plans from $21 a month

Basic is $21/month and Standard $47/month; self-hosted image models cost $0.0047 per image from that allowance. Open Source Unlimited is $149/month with no per-image charge on self-hosted models. Every plan covers Flux, SDXL and 10,000+ community models. Cancel anytime.

100% refund policy on monthly & yearly plans — cancel anytime

Contact Sales

Best Value

Open Source Unlimited

Mission-Critical

$149/month

100% refund policy · cancel anytime

Unlimited Open Source Models

100% refund policy

24x7 Support

15 parallel generations

Access to all APIs

Unlimited generations on all open-source models

For mission critical workloads

Add Team Members

Priority GPU Clusters

Standard

Production

$47/month

100% refund policy · cancel anytime

Moderate Traffic

100% refund policy

Priority Developer Support

10 concurrent API requests

For Production workloads

API access to all models

Prototype

Basic

Prototype

$21/month

100% refund policy · cancel anytime

Moderate Traffic

100% refund policy

Developer Support via Discord/Email

5 concurrent API requests

API access to all models

Shared GPU

Get Expert Support in Seconds

We're Here to Help.

Want to know more? You can email us anytime at support@modelslab.com

View Docs

Yes — Kling 3.0 is available through ModelsLab via REST. You do not need a separate Kuaishou or Kling account; a single ModelsLab API key covers Kling text-to-video, image-to-video and motion-control endpoints.

POST https://modelslab.com/api/v7/video-fusion/text-to-video with `model_id` set to `kling-v3-t2v` and a `prompt`, or POST /api/v7/video-fusion/image-to-video with `model_id` `kling-v3-i2v` and `init_image` as an array of image URLs. Send your API key in the `key` header or body. The response carries an id: POST /api/v7/video-fusion/fetch/{id} for the MP4 URL, or pass a `webhook` URL.

$0.084 per second of video in standard mode without audio, so a 5-second clip costs $0.42. Pro mode multiplies the rate by 1.34, 4k mode by 5.04 and `generate_audio` by 1.5. Kling is a third-party model: on Basic ($21/month) and Standard ($47/month) it is paid from the usage included in the plan, then from wallet balance, and on Open Source Unlimited from wallet balance. Rates checked September 23 2026.

Yes. POST /api/v7/video-fusion/motion-control with `model_id` `kling-v3-motion-control` ($0.126 per second), `init_image` (the character), `init_video` (the movement to copy) and `character_orientation` set to image or video. Optional: `mode` (std or pro) and `keep_original_sound`.

Yes. Kling, Veo, Seedance, Wan and MiniMax all sit under /api/v7/video-fusion/ on the same key. Switch `model_id` to compare quality, latency and cost on the same prompt.

Usually one to a few minutes per clip, depending on duration, mode and the provider queue. Pass `webhook` in the request to be notified when the clip is ready instead of polling fetch/{id}.

ModelsLab is a paid service. Video plans start at $21/month for 3,250 API calls, and Kling, a third-party model, is paid per call from that included usage first, then from wallet balance. Creating an account costs nothing; generating video requires an active plan.

Yes, in a second call. Generate the clip with Kling, then POST /api/v7/video-fusion/lip-sync with `model_id` `lipsync-2`, the clip URL as `init_video` and your speech track as `init_audio`.

`duration` in seconds (the price scales linearly with it), `aspect_ratio`, `mode` (std, pro or 4k) and `generate_audio`. Mode and audio change the per-second rate; see the price answer above.

Kuaishou exposes Kling through its own platform with a separate account and billing relationship. ModelsLab gives you Kling 3.0 through the same REST API as Veo, Seedance and Runway: one key, one bill.

Kling 3.0 is the newest generation of Kuaishou's Kling video models. On ModelsLab it is three model_ids on one key, with no Kuaishou account: kling-v3-t2v (text to video), kling-v3-i2v (image to video) and kling-v3-motion-control (copies the movement of a reference video onto your image).

Kling 3.0 accepts duration as a whole number of seconds from 3 to 15, and the price scales linearly with it: 5 seconds in standard mode is $0.42. Older Kling models snap duration to 5, 10 or 15 seconds. For longer videos, generate clips at scene boundaries and join them.