Get started
Video model

Kling AI: a video generation model, the Chinese answer to Sora

Kling models by Kuaishou in GPTunneL: video from text and from a photo, natural motion, sound in the same render. Pay per second, no subscription.

What the Kling family is good at

Traits Kuaishou keeps from one generation to the next — you can count on them whichever model you pick.

Video from text and from a frame

The model works both ways: it builds a scene from a description or brings a ready picture to life — a photo, a concept, a still from a shoot.

Natural motion

The main reason the family is picked: people, fabric and water move believably and the camera glides without jerks or melting detail.

Sound together with the picture

From 2.6 onwards the model generates speech, ambience and effects in the same render as the footage — no separate audio pass needed.

5 and 10 second clips

A standard clip is 5 or 10 seconds at 720p or 1080p. That is enough for a reel scene, an ad insert or a storyboard frame.

Which model to pick

Kling comes in several tiers — for different jobs and budgets.

v3 MotionMotion on a budget2.5 TurboThe default pick2.6Scenes with soundv3Maximum quality
Best forAnimating a frame, simple motion, trying ideasSocial clips and ad insertsScenes with speech, ambience and effectsComplex scenes, a cinematic frame
Resolution720p and 1080p1080p1080p720p and 1080p
Motion detailMediumHighHighMaximum
SpeedHighestHighMediumBelow average
Price per secondLowestLowModerateHighest

Kling is billed per second of the finished clip and depends on the resolution. Exact per-model prices are on the Pricing

Every Kling model

Release dates follow Kuaishou's official announcements.

2026

Kling v3Kling v3 Motion

The current generation: a clear step up in detail and in the physics of motion. Alongside it runs the lighter v3 Motion — the same motion engine at a fraction of the price per second.

January 2026

Kling 2.6

The generation that learned to generate sound together with the picture — speech, ambience and effects arrive in the same render as the footage.

September 2025

Kling 2.5 Turbo

A faster take on the second generation — the same frame quality noticeably quicker and cheaper. It has been the workhorse of the family ever since.

May 2025

Kling 2.1 MasterKling 2.1

A Master mode with higher detail arrived, along with a choice between 720p and 1080p right in the generation settings.

April 2025

Kling 2.0

The generation that got the model seriously compared with Western video networks — it follows the scene description more closely and holds the motion to the end of the clip.

Full release history
December 2024

Kling 1.6

A clear step forward in animating a picture: the model stopped melting faces and objects on longer shots.

September 2024

Kling 1.5

The first major update: higher resolution and cleaner camera movement.

June 2024

Kling 1.0

Kuaishou's first public model and one of the first video generators to show believable motion physics.

How billing works

There is no subscription: you top up one balance and spend it on any model on the platform.

Pay per second of video

You are charged for the length of the finished clip and its resolution. Not using it costs nothing: no limits, no monthly fee.

One balance for every model

Kling, ChatGPT, Claude, image and music generation — all out of the same wallet. No separate subscription per service.

Top up the way you prefer

International cards, Apple Pay, Google Pay or crypto. The minimum top-up is $5.

See pricing

A sample generation

The clip was generated in Kling from an ordinary text description. There is no need for tags or magic words — the more detail you give about the scene, the light and the camera move, the closer the result gets to what you had in mind.

Kling v3, generated from a text description

How to get started with Kling

Sign in to GPTunneL
One account for every model on the platform.
The camera slowly orbits a blossoming tree covered in frost, a snowy garden around it, green grass beneath
5 seconds1080pKling v3

Sign in whichever way suits you.

The model switches right inside the input — you can change it between clips.

The finished video arrives at the resolution you picked.

Try Kling in GPTunneL

Signing up takes a minute, and $5 on the balance is enough to see whether the model fits your task.

Frequently asked questions

No. Requests go through GPTunneL's infrastructure, so the models open like any ordinary website — no VPN and no proxy.

Kling is a family of video models from the Chinese company Kuaishou, one of the largest video services in China. The model builds a short clip from a text description or brings a ready frame to life: a photograph, a concept, a product shot.

The family is picked for its motion. People, fabric, water and smoke move believably in Kling, and the camera glides — no jerks, no melting detail of the kind other generators trip over. From version 2.6 the model also makes sound: speech, ambience and effects arrive in the same render as the picture. One generation is 5 or 10 seconds at 720p or 1080p; reels, ad inserts and storyboards are assembled from pieces like these. The model understands scene descriptions in dozens of languages, so there is nothing to translate.

Kling models are available in GPTunneL — worldwide, without a VPN, with one balance and billing per second of finished video.