Video

Kling 3.0 API

Kling API on APIMart gives you Kuaishou's Kling 3.0 video model (model ID kling-v3): text-to-video and image-to-video with first and last frames, 3–15 second clips in 720p, 1080p or 4K, optional native audio and multi-shot prompts, all with one APIMart key.

Text-to-VideoImage-to-VideoNative Audio

Get API Key API Docs

Service Highlights

Official Discounts
Pay-as-you-go
Fast & Low Latency

Input

Model

Prompt

14 characters

Mode

Aspect Ratio

Duration5s

Watermark

Audio

Negative Prompt

Multi Shot

Multi Prompt

Add a shot to define segment prompts and durations.

Reference Image

Content Moderation

Output

Video link valid for 72 hours

Kling 3.0 API Pricing

Transparent pricing with no hidden fees. Pay only for what you use.

Variant	Tier	GoldCurrent price	Platinum	Diamond	Official	Your savings
`kling-v3`	720P	0.672Credits(~$0.0672)/sec	0.6384Credits(~$0.0638)/sec	0.6048Credits(~$0.0605)/sec	0.84Credits(~$0.084)/sec	20%
	1080P	0.896Credits(~$0.0896)/sec	0.8512Credits(~$0.0851)/sec	0.8064Credits(~$0.0806)/sec	1.12Credits(~$0.112)/sec	20%
	720P+Sound	1.008Credits(~$0.1008)/sec	0.9576Credits(~$0.0958)/sec	0.9072Credits(~$0.0907)/sec	1.26Credits(~$0.126)/sec	20%
	1080P+Sound	1.344Credits(~$0.1344)/sec	1.2768Credits(~$0.1277)/sec	1.2096Credits(~$0.121)/sec	1.68Credits(~$0.168)/sec	20%
	4K	4.2856Credits(~$0.4286)/sec	4.0713Credits(~$0.4071)/sec	3.857Credits(~$0.3857)/sec	5.357Credits(~$0.5357)/sec	20%
	4K+Sound	4.2856Credits(~$0.4286)/sec	4.0713Credits(~$0.4071)/sec	3.857Credits(~$0.3857)/sec	5.357Credits(~$0.5357)/sec	20%

* Actual costs are subject to final output.

Kling API - Kuaishou Kling 3.0 Video Generation

Kling API access to Kuaishou Kling 3.0 on APIMart: turn a prompt or one to two frames into a 3–15 second video in 720p, 1080p or 4K, with optional audio, multi-shot storyboards and per-second billing.

Get API Key

50K+

Active Users

99.9%

Uptime

Faster

—

Cost Savings

Key Features of the Kling 3.0 API

What the kling-v3 model can do, mapped to the request parameters you send

Text to Video with the Kling 3.0 API

Describe the scene, subject, action and camera in prompt and the Kling 3.0 API returns a finished clip. No image is needed: when image_urls is omitted the request runs as text to video.

Image to Video with First and Last Frames in the Kling V3 API

Pass one image in image_urls to use it as the first frame, or two images to set the first and last frames. The Kling API animates everything in between; the output ratio may follow the input image.

Native Audio in Kling Video 3.0

Set audio to true and the Kling 3.0 API generates the clip with sound in the same task. Audio is billed as its own per-second tier, so leave it off for silent footage you will score later.

Kling 3.0 Pro Mode: 720p, 1080p and 4K

In the Kling API, mode picks the output tier: std for 720p, pro for 1080p and 4k for 4K. Draft in std, then render the takes you keep in pro or 4K.

3–15 Second Clips with the Kling 3 API

Kling 3.0 API duration accepts any whole number from 3 to 15 seconds (default 5). Since billing is per second, shorter test clips keep iteration cheap.

Multi-Shot Storytelling with Kling AI 3.0

Turn on multi_shot and describe up to 6 shots in multi_prompt, each with its own prompt and length. Shot durations must add up to the total duration, so one Kling API request returns a cut sequence.

Consistent Characters with Kling 3.0 Elements

element_list defines up to 3 subjects, each from 2–4 reference images, and the prompt calls them with @name. In the Kling 3.0 API it must be used with at least one image in image_urls.

Aspect Ratios and Negative Prompts in the Kling AI API

Choose 16:9, 9:16 or 1:1, list what to avoid in negative_prompt, and add optional watermark or nsfw_check (content review before submission) settings.

What You Can Build with the Kling Video API

Typical Kling 3.0 projects and the settings that fit them

Vertical Social Ads with Kling 3.0

Generate 9:16 Kling API clips with audio for Reels, Shorts and TikTok, test several hooks in std, then re-render the winner in pro.

Product Videos from Catalog Photos

Use a packshot as the first frame and let the Kling API add camera moves and motion, keeping the product as it appears in your photo.

Storyboards and Short Narratives

Plan a 15-second, multi-shot sequence in one Kling 3.0 API request to preview pacing and framing before a full production.

Scene Transitions with First and Last Frames

Give the Kling 3.0 model a start and an end image to create controlled transitions, before-and-after reveals or loops.

Education & Explainers

Use the Kling API to turn concepts into short visual explanations, with narration-style audio when the audio option is on.

Video Features in Your Own App

Add text-to-video or photo animation to a SaaS product through the asynchronous Kling API: submit, poll, store the result URL.

How to Get a Kling API Key and Start Building

From sign-up to your first Kling 3.0 video in three steps

Create Your Kling API Key

→

Test Kling 3.0 in the Playground

Try prompts, frames, modes and durations above and check the cost estimate before you write code.

→

Integrate the Kling 3.0 API

Follow the API reference to submit tasks from your backend and poll for results.

Start Creating Videos

Kling 3.0 API Example: Text to Video and Image to Video

Call the Kling API on APIMart with one POST request. Generation is asynchronous: submit a task, then poll the task endpoint to get the video URL.

1
Authenticate Kling API requests
Save the key you created above in the APIMART_API_KEY environment variable and send it as Authorization: Bearer <key> on every Kling API request.
2
Submit a Kling 3.0 task
POST /v1/videos/generations with model kling-v3, your prompt, mode, duration and aspect_ratio. Add image_urls for image to video. The response contains the task ID in data[0].task_id.
3
Get the Kling 3.0 video result
Poll GET /v1/tasks/{task_id} until status is completed, then read the video URL from data.result.videos[0].url[0]. Download it before expires_at, because result links expire.
See the Kling 3.0 API reference for every parameter, scenario and response field.

POST /v1/videos/generations

curl -X POST https://api.apimart.ai/v1/videos/generations \
  -H "Authorization: Bearer $APIMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kling-v3",
    "prompt": "A golden retriever runs along a beach at sunset, slow motion, low tracking shot, cinematic lighting",
    "mode": "std",
    "duration": 5,
    "aspect_ratio": "16:9"
  }'

import os
import time
import requests

API_BASE = "https://api.apimart.ai"
headers = {"Authorization": f"Bearer {os.environ['APIMART_API_KEY']}"}

# 1. Submit the generation task
payload = {
  "model": "kling-v3",
  "prompt": "A golden retriever runs along a beach at sunset, slow motion, low tracking shot, cinematic lighting",
  "mode": "std",
  "duration": 5,
  "aspect_ratio": "16:9"
}
resp = requests.post(f"{API_BASE}/v1/videos/generations", headers=headers, json=payload, timeout=60)
resp.raise_for_status()
task_id = resp.json()["data"][0]["task_id"]

# 2. Poll the task until it finishes
for _ in range(200):
    task_resp = requests.get(f"{API_BASE}/v1/tasks/{task_id}", headers=headers, timeout=30)
    task_resp.raise_for_status()
    task = task_resp.json()["data"]
    if task["status"] == "completed":
        print(task["result"]["videos"][0]["url"][0])
        break
    if task["status"] == "failed":
        raise RuntimeError(task.get("error", {}).get("message", "generation failed"))
    time.sleep(10)
else:
    raise TimeoutError(f"task {task_id} did not finish in time")

curl -X POST https://api.apimart.ai/v1/videos/generations \
  -H "Authorization: Bearer $APIMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kling-v3",
    "prompt": "Smooth cinematic transition from the first frame to the last frame, gentle camera push-in",
    "image_urls": [
      "https://example.com/frame-start.jpg",
      "https://example.com/frame-end.jpg"
    ],
    "mode": "std",
    "duration": 5
  }'

TASK_ID="task_xxx"  # replace with the task_id returned by the POST request
curl "https://api.apimart.ai/v1/tasks/$TASK_ID" \
  -H "Authorization: Bearer $APIMART_API_KEY"

Kling 3.0 API Parameters

Parameter	Type	Required	Default	Description
`model`	string	Required	Default: kling-v3	Use kling-v3 for Kling 3.0.
`prompt`	string	Required	Default: —	Kling 3.0 API scene, action and style description. English prompts are recommended. Can be omitted when multi_shot is true.
`negative_prompt`	string	Optional	Default: —	Content to exclude, for example "blurry, low quality, distorted".
`mode`	string	Optional	Default: std	std (720p), pro (1080p) or 4k.
`duration`	integer	Optional	Default: 5	Kling 3.0 API clip length from 3 to 15 seconds. Send a number, not a quoted string.
`aspect_ratio`	string	Optional	Default: 16:9	16:9, 9:16 or 1:1. With image to video, the input image may override it.
`image_urls`	array	Optional	Default: —	Up to 2 public image URLs for Kling 3.0 API image to video: 1 image is the first frame, 2 images are the first and last frames.
`audio`	boolean	Optional	Default: false	Generate the video with sound.
`watermark`	boolean	Optional	Default: —	Add a watermark to the output.
`nsfw_check`	boolean	Optional	Default: false	Review the prompt and input images with omni-moderation-latest before submission.
`multi_shot`	boolean	Optional	Default: false	Enable Kling 3.0 API multi-shot mode. shot_type is then required.
`shot_type`	string	Optional	Default: —	customize or intelligence. Required when multi_shot is true.
`multi_prompt`	array	Optional	Default: —	1–6 shots, each with index, prompt (up to 512 characters) and duration (at least 1 second). Durations must add up to duration. Required when shot_type is customize.
`element_list`	array	Optional	Default: —	Up to 3 subjects, each with name, description and 2–4 element_input_urls; reference them in the prompt as @name. Requires at least 1 image in image_urls.

User Reviews for Kling 3.0

Real feedback from Kling 3.0 users worldwide. These reviews describe experiences across the Kling V3 series.

“The cinematic quality is incredible! The 15-second duration option gives us much more creative freedom for storytelling.”

Sarah Johnson

Creative Director

“As a developer, the unified Omni API makes integration a breeze. One model handles all our multi-modal generation needs.”

James Liu

Senior Developer

Kling API Model Family

Compare Kling video models on APIMart. Check each model’s availability, capabilities and pricing on its own page.

Model	Best for	Resolution	Duration
Kling 4.0 API	Text, frame and multimodal generation, video editing and extension, always with stereo audio	720P / 1080P	3–30 s
Kling 3.0 APIThis page	Cinematic text-to-video and image-to-video with first/last-frame control and multi-shot prompts	720P / 1080P / 4K	3–15 s
Kling 3.0 Omni API	All-in-one generation from text, images and video, with reference roles and video editing	720P / 1080P / 4K	3–15 s
Kling 3.0 Turbo API	Speed-optimized text-to-video and first-frame image-to-video	720P / 1080P	3–15 s
Kling Video O1 API	Thinking-driven generation focused on visual fidelity	720P / 1080P	5 s / 10 s
Kling 2.6 API	Native audio (pro mode), dynamic masks and camera control	720P / 1080P	5 s / 10 s
Kling Motion Control 2.6 API	Transfer motion from a reference video onto the subject of your image	720P / 1080P	Follows reference video
Kling 3.0 Motion Control API	Newer V3 generation of reference-video motion transfer	720P / 1080P	Follows reference video

Kling 3.0 API FAQ

Common questions about Kling 3.0 models, pricing and API access

What is Kling 3.0 (Kling Video 3.0)?

Kling 3.0 is Kuaishou's third-generation AI video model, released as Video 3.0 on the Kling AI platform. On APIMart it is called with the model ID kling-v3 and generates 3–15 second clips from text or from one or two frames, in 720p, 1080p or 4K, with optional audio.

When was Kling 3.0 released?

Kuaishou announced Kling 3.0 in early February 2026, together with Kling 3.0 Omni and the Kling Image 3.0 models. Access started with Kling AI subscribers and was later extended to all users. The kling-v3 model is available through the APIMart Kling API now.

What is the Kling API and how does it work?

The Kling API is programmatic access to Kling video models. On APIMart you send one POST request to /v1/videos/generations with model kling-v3; the response returns a task ID, and you poll the task until the video URL is ready. The example section on this page shows the full flow.

How is Kling 3.0 different from Kling 2.6?

Kling 3.0 accepts any length from 3 to 15 seconds and adds a 4K tier, multi-shot prompts and reusable subjects (element_list). On APIMart, Kling 2.6 offers 5 or 10 second clips at 720p or 1080p, with native audio in pro mode, dynamic masks and camera control.

Kling 3.0 vs Omni vs Motion Control: which Kling API should I use?

Use Kling 3.0 (this page) when you start from a prompt or one to two key frames. Choose Kling 3.0 Omni to give images roles, reference the features of a video or edit an existing video. Choose Kling 3.0 Motion Control to copy the movement of a reference video onto the subject of your image. Choose Kling 3.0 Turbo when speed matters more than extras. The Kling API Model Family table on this page links to each model.

What parameters does the Kling 3.0 API support?

Required: model (kling-v3) and prompt; prompt can be omitted when multi_shot is on. Optional: mode (std, pro, 4k), duration (3–15, default 5), aspect_ratio (16:9, 9:16, 1:1), image_urls (up to 2), negative_prompt, audio, watermark, nsfw_check, multi_shot with shot_type and multi_prompt, and element_list.

Can the Kling 3.0 API do image to video?

Yes. Add image_urls to a Kling 3.0 API request: one image becomes the first frame, two images become the first and last frames. Images must be public URLs without hotlink protection, and the output aspect ratio may follow the image instead of aspect_ratio.

How do audio, resolution and duration affect Kling 3.0 API pricing?

Kling 3.0 API cost = the per-second rate of your tier × duration. The tier is set by mode (720p std, 1080p pro or 4K) and by whether audio is on; at 720p and 1080p, audio has a higher rate. For example, a 10-second 1080p clip with audio costs 10 × the 1080P+Sound rate in the pricing table on this page.

How do I call the Kling API, and where are the docs?

Submit POST /v1/videos/generations, read data[0].task_id, then poll GET /v1/tasks/{task_id} until status is completed and read data.result.videos[0].url[0]. The full Kling 3.0 API reference, with every parameter and scenario, is on docs.apimart.ai.

How do I get a Kling API key?

Sign up for APIMart, open the API Keys page in your dashboard and create a key. The same key works for every Kling model and every other model on the platform, with pay-as-you-go billing and no subscription.

Is the Kling API free?

No. There is no free quota. It is pay-as-you-go, you only pay for successful requests, and all prices are 20% off for members.

Is there a discount on Kling API pricing?

Yes. Kling API calls are billed per second of generated video. The pricing table on this page lists all available options. Prices shown are Gold member prices (20% off); Platinum and Diamond members save even more, up to 28%. You only pay for successful requests.

Can I use Kling 3.0 videos commercially?

Yes, videos generated through the Kling API on APIMart can be used in commercial projects, subject to the APIMart terms of service and Kling's usage policies.

Is Kling V3 the same as Kling 3.0 and Kling AI 3.0?

Yes. Kling V3, Kling AI 3.0, Kling Video 3.0, Kling 3 and Kling3.0 all refer to the same Kuaishou model, which APIMart exposes as kling-v3. Kling 3.0 Omni, Turbo and Motion Control are separate models with their own pages.

Related Models

Explore more models in the same category.

View All Models →