Create Task

This is provided as part of the Kling API from PiAPI.
This service allows you access to APIs for the most advanced video generation model - Kuaishou's KlingAI.
You can create Text-to-video, Image-to-video, Extend video, Lipsync task and Effects task through this endpoint.

INFO

Text-to-video vs Image-to-video

If the parameter image_url is left null then the task would be a text-to-video task; if filled out, then the task would be an image-to-video task.

About version 1.5 1.6 2.0 2.1 and 2.1-master

The V1.5/1.6 model through the API currently has all the features and constraints as per klingai.com
pricing for 2.1 is the same as 1.6
pricing for 2.0/2.1-master: 2.0-pro-5s: USD0.96/generation, 2.0-pro-10s: USD1.92/generation

Video Lengths

For the video lengths (5s vs 10s) of the API-generated-videos for each model (1.0 1.5 1.6 2.0 2.1 and 2.1-master) and modes (std vs pro) combination, please refer to what is allowed on klingai.com; the API currently aligns with the website.

Request

Header Params

x-api-key

string

required

Your API Key used for request authorization

Body Params application/json

model

enum<string>

required

the model name

Allowed value:

kling

task_type

enum<string>

required

type of the task. If you want to extend video,set to extend_video and include origin_task_id filed in input; if you want lipsync, use lip_sync and include origin_task_id in input

Allowed values:

video_generationextend_videolip_synceffects

input

object

required

the input param of the kling task, depends on the task_type .refer to the example for more details.

prompt

string

optional

<= 2500 characters

negative_prompt

string

optional

<= 2500 characters

cfg_scale

string

optional

float between 0 and 1, it is strongly recommended to use 0.5

duration

enum<integer>

optional

Allowed values:

510

aspect_ratio

enum<string>

optional

only required in text-to-video task

Allowed values:

16:99:161:1

camera_control

object

optional

mode

enum<string>

optional

Allowed values:

stdpro

version

enum<string>

optional

the model version

Allowed values:

1.01.51.62.02.12.1-master

image_url

string

optional

Only required in image-to-video task. Initial frame of the video.No larger than 10MB, and each side must be greater than 300 pixels.

image_tail_url

string

optional

End frame of the video. No larger than 10MB, and each side must be greater than 300 pixels.Need to be used with image_url.

origin_task_id

string

optional

The task_id of the video to extend or lip_sync.

tts_text

string

optional

The text that you want to lipsync in the video, used in lip_sync task only; will be ignored if local_dubbing_url is set

tts_speed

number

optional

The speed of the lip_sync speech, used in lip_sync task with valid tts_textonly.

>= 0.8<= 2

Default:

tts_timbre

string

optional

The voice that you want to use in lip_sync task with valid ttx_text, the full voice name list and demo list here: https://klingai.com/api/lip/sync/ttsList?type=

local_dubbing_url

string <uri>

optional

Use your audio file for lip_sync through this parameter. All tts related param will be ignored if this is set. Audio format:mp3``wav``flac``ogg, should be less than 20 MB and shorter than 60 seconds. After the audio is submitted, we will trim it to the exact length of the original video.

Example:

https://example.com/hello.mp3

motion_brush

object (motion brush)

optional

It's highly recommended that you read our example doc before acutally use motion brush

effect

enum<string>

optional

your choice of kling effect, see Kling Effects

Allowed values:

squishexpansion

config

object

optional

webhook_config

object

optional

service_mode

enum<string>

optional

This field allows for more flexible switching between HYA and PAYG.

public->Pay As You Go(PAYG)

private->Host Your Account(HYA)

Allowed values:

publicprivate

Example

{
    "model": "kling",
    "task_type": "video_generation",
    "input": {
        "prompt": "White egrets fly over the vast paddy fields",
        "negative_prompt": "",
        "cfg_scale": 0.5,
        "duration": 5,
        "aspect_ratio": "1:1",
        "camera_control": {
            "type": "simple",
            "config": {
                "horizontal": 0,
                "vertical": 0,
                "pan": -10,
                "tilt": 0,
                "roll": 0,
                "zoom": 0
            }
        },
        "mode": "std"
    },
    "config": {
        "service_mode": "",
        "webhook_config": {
            "endpoint": "",
            "secret": ""
        }
    }
}

Request samples

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

Responses

🟢200Success

application/json

Body

code

integer

required

data

object

required

task_id

string

required

model

string

required

task_type

string

required

status

enum<string>

required

Hover on the "Completed" option and you coult see the explaintion of all status: completed/processing/pending/failed/staged

Allowed values:

CompletedProcessingPendingFailedStaged

input

object

required

output

object

required

Text-to-video vs Image-to-video#

About version 1.5 1.6 2.0 2.1 and 2.1-master#

Video Lengths#

Request

Request samples

Responses

Text-to-video vs Image-to-video

About version 1.5 1.6 2.0 2.1 and 2.1-master

Video Lengths