China's new text-video model “Kling” gives tough competition to open AI's Sora

Kling is a new text-to-video conversion model launched by Kuaishou Technology, a Chinese company known for its TikTok-like short-video platform.

New DelhiWhile OpenAI's text-to-video conversion model is not even available to users yet, new videos and new video generation AI models are already making waves, providing tough competition for ChatGPT makers. Kling is a new text-to-video conversion model launched by Kuaishou Technology, a Chinese company known for its TikTok-like short video platform.

advertisement

advertisement

Kling, an innovative text-to-video AI model, can create highly realistic videos from simple text prompts. The arrival of the Kling platform could give Open AI's Sora a much-needed competitor.

What can “Kling” do?

Kling uses similar technology to Sora to produce 1080p high definition video with up to two minutes of smooth 30 frames per second animation. Kling's deeper understanding of the physical world allows it to simulate complex, realistic movements.

Like Sora, Kling utilizes the Diffusion Transformer, a technology that reads a prompt and converts it into a visual from a single frame. Given a prompt, Kling can create a video of a panda playing the guitar.

The Chinese company released a video demonstrating the capabilities of its new text-to-video conversion model, with some of the videos generated by Kling outperforming the output of Open AI's Sora.

The Kling model also gives users freedom when it comes to aspect ratios.The proprietary Large Language Model (LLM) uses advanced 3D face and body reconstruction backed by the company's proprietary technology 3D VAE, which enhances various distinctive aspects of the video, including realistic facial expressions and improved limb movements.

Kling: Race to the West

The launch of Kling marks a major breakthrough in the field of artificial intelligence and a shift in power to the East. Kling offers longer video duration, better resolution, and more accurate results, outperforming OpenAI's Sora on many metrics.

Moreover, Kling is not the only text video model to come out of China: in April 2024, Vidu AI was released in China. These latest developments in the eastern part of the world will undoubtedly challenge the dominance of Western companies.

But China has a history of facing restrictions in many countries, which could make it difficult for Chinese products to take hold globally.