Kling AI

Lack of transparency

Kling AI does not disclose detailed information about model architecture and inference infrastructure. Thus, the environmental impacts are estimated with a very low precision.

This guide focuses on estimating video generation impacts for Kling AI models with the video_impacts function.

Video Generation

Supported Kling AI video models:

Model	Identifier
Kling 1.6	klingai/kling-v1.6
Kling 2.6	klingai/kling-v2.6
Kling 3	klingai/kling-v3

from ecologits.estimations import video_impacts

model = "klingai/kling-v3"
resolution = "1080p"
duration = 6

impacts = video_impacts(
    model_name=model,
    resolution=resolution,
    duration=duration,
)

print(impacts.energy.value)
print(impacts.gwp.value)

Set with_audio=False when your request generates video without audio:

impacts = video_impacts(
    model_name=model,
    resolution=resolution,
    duration=duration,
    with_audio=False,
)