Kling AI, a cutting-edge tool for generating imaginative images and videos using state-of-the-art generative AI techniques, held its “From Vision to Screen” Model 2.0 Launch Event in Beijing on April 15, announcing an upgrade to its foundational model and officially unveiling the KLING 2.0 Video Generation Model and KOLORS 2.0 Image Generation Model globally.

Redefining human-AI interaction

Since its launch last June, Kling AI has undergone more than 20 iterations. As of now, its global user scale has surpassed 22 million. Over 15,000 developers from around the world have integrated Kling’s Application Programming Interface (API) into a wide range of industry applications.

“Kling AI has consistently focused on improving the foundational capabilities of its models, enhancing image quality, and introducing innovative features to meet the diverse needs of its users,” Mr. Gai Kun, Senior Vice President of Kuaishou and Head of the Community Science Department, said at the event, adding that the tool aims to empower everyone to tell compelling stories with AI, enabling more precise and complex creative expression.

With the latest upgrade to its foundational models, the KLING 2.0 model continues to lead globally in areas such as dynamics, prompt adherence, and visual aesthetics, while the KOLORS 2.0 model has seen improvements in prompt adherence, cinematic visual quality, and representation of artistic styles.

Gai said that both models have ranked first in the industry based on the team’s internal multi-metric comparative evaluation.

According to the latest ranking of video generation models validated by the global AI benchmark organization Artificial Analysis on March 27, Kuaishou’s Kling 1.6 Pro (High-Quality Mode) claimed the top spot in the Image-to-Video category with a benchmark Arena ELO score over 1000, followed by Google’s Veo 2 and Pika Art, which ranked second and third, respectively.

Gai said that AI holds tremendous potential for supporting creative expression; however, the industry still fails to meet user demands. To truly realize the vision of “telling great stories with AI,” it is essential to comprehensively enhance foundational model capabilities and define a new language for human-AI interaction.

In this 2.0 model iteration, Kling AI officially introduces a new interactive concept for AI video generation: Multi-modal Visual Language (MVL). This concept enables users to efficiently convey complex, multi-dimensional creative ideas—such as identity, appearance, style, scenes, actions, expressions, and camera movements—directly to AI by integrating multi-modal information like image references and video clips.

Multi-modal features

Based on the new MVL concept, the brand-new Multi-Elements Editor and Image Editing features were unveiled by Mr. Zhang Di, Vice President of Kuaishou and Head of Kling AI, at the launch event.

For instance, based on the existing video, the Multi-Elements Editor allows users to swap, add, or delete elements from the video with text or image inputs, empowering creators with more creative freedom and flexibility in editing.

According to Zhang, image-to-video generation currently accounts for about 85 percent of Kling AI’s video creation, with image quality playing a crucial role in the video generation process. In the field of image generation models, Kuaishou’s Kolors2.0 also leads the industry.

He said that KOLORS 2.0’s text-to-image capabilities have undergone a comprehensive upgrade, including a significant improvement in prompt adherence, a substantial enhancement in cinematic aesthetic expression, and support for over 60 stylizations for image transformation. As a result, the model’s creativity and imagination in generating images have greatly improved.

Comments URL: https://news.ycombinator.com/item?id=43724227

Points: 1

# Comments: 1

Categories: Hacker News

Biographical Information Summary - This is Just a Summary Joe Pearce
About Joe Pearce joeintenn
Links Joe Pearce
Flounder's Keylime Pie is the Best in the World, At Least I Think So... Joe Pearce
Harley Ride Joe Pearce
Cobra with New Cover Joe Pearce
Mustang Cobra After Ceramic Coating Joe Pearce
Carter County Cruise In Joe Pearce
2003 Ford Mustang SVT Cobra Convertible NAPA Auto Car Show Top 10 Joe Pearce
Ponies in the Smokies - Mustang Trophy Joe Pearce

Hacker News

Human-AI collaboration is not collaborative yet

PIM-LLM: A High-Throughput Hybrid PIM Architecture for 1-Bit LLMs

Tesla speeds up odometers to avoid warranty repairs, US. lawsuit claims

A 1980s toy robot arm inspired modern robotics

Trump says U.S. to sign rare earths minerals deal with Ukraine next week

Fatima Hassouna and 9 of her family members murdered

macOS Finder "Connect to Server" accepts ftps URIs but uses plain FTP

Honest and Elitist Thoughts on Why Computers Were More Fun Before

Four charts sum up the state of AI and energy

Building UK Defence Innovation

Ava: The Art and Science of Image Discovery at Netflix

Tristan Harris Congress Testimony: Understanding Persuasive Technology [video]

Choosing the right Python task queue

Ranked: Countries Investing the Most in R&D

Remi Bader Chose Herself

Microsoft's Blue Screen of Death May Get a Gloomier Update

Intuit, Owner of TurboTax, Wins Battle Against America's Taxpayers

Wall Street (Kinda) Is Main Street

Students at Military Base Schools Walk Out to Protest Trump Policies

Kling AI 2.0 Launches with Multimodal Video/Image Editing

Pages

Welcome to Joe Pearce's Home Page.

Web page offered by Joe Pearce © 2004 - 2025 - All rights reserved.

Thanks to the ETSU Computer and Information Sciences Department.

Thanks to the NSTCC Computer and Information Sciences and Computer Engineering Technologies Department.

This is my Favicon.

You are here