Kuaishou’s Text-To-Video 'Kling AI' Goes Global To Compete Against OpenAI, And More

The AI industry is trending more than ever, thanks to the rise of Large Language Model-powered generative AIs.

Soon after OpenAI popularized this with the introduction of OpenAI, two could play this game. Then three, four, five, and more followed, each developing their own products and solutions.

While most products come from the West, the East isn't falling behind.

China knows well how lucrative this generative AI trend is, and one of the leading tech companies in the country has debuted a text-to-video generative AI to compete directly against the likes of OpenAI Sora, Luma AI's Dream Machine and more.

Kuaishou has what it calls the Kling AI, which quickly goes viral after its debut, and quickly gained traction after its global debut.

Back in July, Kuaishou showcased several fresh features for its text-to-video model Kling AI at the World Artificial Intelligence Conference (WAIC) in Shanghai, including the ability to generate videos up to 10 seconds.

The AI was introduced to only a select number of users through invitations.

Those early users tested the AI, and soon, the clips they created were uploaded to social media, as many were astonished by the results.

More than 500,000 users have applied to help beta test Kling, revealed Gai Kun, Senior Vice President of Kuaishou, at a WAIC forum.

And sooner than later, millions of videos have already been generated.

As a result, AI-generated videos by Kling AI quickly flooded the Chinese internet, with countless clips featuring characters from historical films performing modern-day tasks, spawning multiple memes.

Kuaishou provided practical tips, advising users to use simple words and sentence structures and to avoid overly complex language.

Regardless, Kuaishou has been preparing for Kling AI's global debut.

Aware of how quickly Sora is gaining momentum, Kuaishou hopes that Kling AI can catch up sooner rather than later. By opening the platform to the public, Kuaishou aims to attract more English-speaking users to create AI-generated videos.

At this time, Kling AI only understands Mandarin and English.

Kuaishou has set the web portal for its AI video tool for international users to KlingAI.com, while retaining the original Chinese version at kling.kuaishou.com.

While Kling AI has received mixed reviews, there are instances where its capabilities surpass those of most rivals.

First, Kling AI supports both text-to-video and image-plus-text-to-video generation.

Second, generation times average around 5 minutes per video. For a 5-second clip at 30fps, that's roughly 1 minute per second of video, or half a second per frame.

As for image generation, Kling AI can do this almost instantaneously, taking just a few seconds per image. This speed is impressive, considering that traditional AI image generators often require several seconds to produce a single high-resolution image.

Third, Kling AI offers versatility in its output.

For instance, users can choose between three video resolutions: a 960x960 square format, 720p landscape, and 720p portrait mode.

Additionally, the platform boasts a camera control feature, allowing users to select from static, horizontal, vertical, pan, tilt, or rolling camera movements.

Fourth, users can fine-tune their creations by adjusting the balance between creativity and relevance. Favoring relevance ensures closer adherence to the prompt, while prioritizing creativity can result in more imaginative and visually striking outputs. The platform also supports negative prompts, potentially enhancing accuracy.

Following its global rollout, Kuaishou has also developed a business model for Kling AI.

Users who wish to use Kling AI for free can do so, but are limited to just a few credits per day.

Those who want to lift this limitation must subscribe to a paid program, which increases their daily generation credits.

While Kling AI can create astonishing results, it struggles with generating letters and numbers accurately.

Like most generative AI products, Kling AI isn’t particularly sensitive to numbers. For example, a prompt like “10 puppies on the beach” may not consistently result in the correct number of puppies in the output.

Kuaishou also imposes certain limitations on the AI, preventing it from generating violent content, depictions of famous people, or NSFW material.

Version 1.0 is just the beginning for Kling AI. Kuaishou has plans to introduce advanced features in future updates, including a high-quality mode for full HD video production, extended generation times, end-frame additions in image-to-video conversions, and more varied camera movements.

Besides Kling AI, Kuaishou has developed several other large model series, including the language model KwaiYii and the image-focused model Kolors.

Kuaishou launched its AI strategy in 2023. According to CEO Cheng Yixiao, generative AI offers a "very rich combination of business scenarios and huge value potential" for the content platform.