AI video enters the 'clash of the titans' moment: ByteDance and Kuaishou spark an industrial revolution. Who will be the new favorite in the capital market?

The field of AI video generation is witnessing a true 'celestial showdown' moment.

Recently, ByteDance released Seedance 2.0, in conjunction with $KUAISHOU-W (01024.HK)$ the iteration of Kuaishou 3.0, pushing the AI video sector from a 'large model arms race' towards 'industrial implementation.' This upgrade sees both giants focusing on enhancing the 'comprehension' ability in video creation, striving for breakthroughs in multimodal integration, camera language, and narrative logic.

Additionally, there are reports that Xiaohongshu's technical team is developing an AI product for video editing called OpenStoryline, which is currently in the testing phase and may be open-sourced later.

Why has this technological upgrade ignited the capital markets? From 'selling shovels' to 'mining gold,' how is the investment logic undergoing dramatic changes? This article will provide fellow investors with an in-depth analysis.

Technical Showdown: Seedance 2.0 VS Kuaishou 3.0

According to Dongwu Securities, Seedance 2.0 shows significant improvement over version 1.5. Seedance 2.0 has further broken through the 'controllability' bottleneck, moving toward industrial-grade applications. Task self-decomposition and high completion rates are the industry’s straightforward evaluation of Seedance 2.0, with almost entirely positive reviews on X. Key features of Seedance 2.0 include:

1) Self-shot composition and self-motion:Automatically plan shots and motion based on descriptions;

2) Comprehensive multimodal references: Up to 9 images, 3 video clips, and 3 audio clips can be provided simultaneously. These references can specify actions, special effects, camera movement styles, character appearances, scene ambiance, and even sound effects;

3) Synchronized audio-visual generation:This is a breakthrough improvement. Seedance 2.0 no longer adds audio as a post-production element but generates it synchronously with the video, producing high-quality matching sound effects and soundtracks while generating the video.

4) Multi-camera narrative capability:Seedance 2.0 can generate coherent narrative videos of 30-60 seconds (breaking the previous limit of 15 seconds), maintaining consistency of characters and scenes across multiple shots.

Keling 3.0 achieves more stable integrated output. Keling 3.0 integrates multimodal input and output, emphasizing 'long videos' and 'controllability.' The 15-second extended generation, Omni model precision control, and 'audio-visual synchronization' features have yielded good results in traditional fields such as gaming, advertising, film, and short films. The main functional highlights of Keling 3.0's four products are as follows:

1) Video 3.0:Narrative and control have evolved deeply, including intelligent storyboarding, intelligent audio-visual synchronization, and precise text;

2) Video 3.0 Omni:All-in-one reference, audio-visual integration. Compared to the O1 version, similarity has significantly improved, with more sensitive and precise responses to complex text instructions.

3) Image 3.0:Cinematic-grade lighting reconstruction that efficiently supports professional needs, enabling creation in single/multi-image compositions. It allows for batch adjustments and unified optimization of multiple images, effectively creating a complete visual system with consistent style across various scenarios.

4) Image 3.0 Omni:4K ultra-clear narrative quality, batch generation, balancing consistency and flexibility.

Which companies are expected to benefit?

Orient Securities believes that video generation has entered the era of precise and controllable 'dashboard.' New-generation domestic models like DreamSeed 2.0 and Kuaishou's KeLing 3.0 generally support multi-modal references such as image/audio/video input. Single-generation duration extends up to 15 seconds, adding intelligent storyboarding, multi-camera storytelling, and secondary editing capabilities—significantly lowering content production barriers for both B-end professional creators and C-end general users.This accelerates AI-driven creation in high-frequency media scenarios like entertainment, short dramas, and advertisements.

In fact, if ByteDance and$KUAISHOU-W (01024.HK)$are the 'shovel sellers' providing infrastructure,then companies with massive IPs and copyright services are the mine owners sitting on their own 'gold mines.'The market is focusing on two core directions: copyright confirmation (infrastructure extension) and IP value reevaluation (asset monetization). The specific key beneficiary areas are as follows:

Track One: Copyright Confirmation – the 'Digital Infrastructure' of the AI Era

$VOBILE GROUP (03738.HK)$ As a globally leading provider of online video content protection services, the proliferation of AI-generated content has made it more necessary for IP owners to use technological means for rights confirmation and protection, which is at the core of this company's business logic. It currently provides exclusive copyright technology services for Red Fruit Short Drama. This company has surged nearly 20% over the past two trading days.

The explosive growth of AI content has created two strong demands – rights confirmation and transactions. Guoyuan International analysts believe that as AI-generated content grows explosively, demand for copyright confirmation and assetization will increase rapidly. As an industry hub, the company is well-positioned to benefit significantly from the acceleration of commercialization.

Guohai Securities previously pointed out that the surge in AI video content presents a golden opportunity for copyright protection. Copyright protection provider Fobor, with its Vobile MAX rights confirmation revenue-sharing platform and DreamMaker creation tool ecosystem, offers an end-to-end solution covering 'creation-rights protection-circulation.' Benefiting from the industry's shift towards an 'AI creation + precise push' model, it stands to become a key beneficiary of copyright infrastructure in the AI content era.

Track Two: IP Asset Reevaluation – From 'Static Inventory' to 'Dynamic Cash Flow'

A. Content Giant: China Literature Group

$CHINA LIT (00772.HK)$The company has also surged over 20% in the past two trading days.The company’s CEO and President, Hou Xiaonan, stated in a Lunar New Year internal letter that,The faster AI develops, the more precious original content becomes, and the value of good stories will be amplified like never before. China Literature’s vast collection of high-quality IPs is set to embrace new opportunities for visualization.Yuewen will fully embrace AI in content incubation, premium content production, IP development, and industry globalization. It is making an all-out effort to deploy new scenarios such as AI comic dramas, enhancing the development efficiency and value potential of IPs.

BOCI Securities noted that, according to Xinhua News Agency data from January 2026, Yuewen Group already has more than ten works with over 100 million views and over 80 comic drama works with over 10 million views. With the release of Seedance 2.0, the production cost of comic dramas has been further reduced while production quality has improved, potentially leading to accelerated industry growth.

B. Business Model Restructuring: Pop Mart

Market observers believe that $POP MART (09992.HK)$ may be the biggest beneficiary of this round of technological iteration. For a long time, IPs like Molly and Labubu have been criticized for having 'an image but no soul.' The 'multi-camera consistency' of AI models resolves the pain points of high cost and high risk in traditional animation, allowing Pop Mart to generate high-quality short films at extremely low costs, injecting narrative depth into IPs and significantly raising their valuation ceiling.

C. Cost Reduction and Efficiency Improvement Model: Disney

For $Disney (DIS.US)$ , AI is a 'double-edged sword' where short-term benefits outweigh drawbacks. Although facing challenges from content oversaturation, given the hundreds of billions of dollars in production costs, AI tools provide immediate cost reductions in pre-production previews, background rendering, and other areas, directly repairing profit margins. Moreover, amid the flood of AI-generated content, emotional connections established by top-tier IPs become even scarcer and more valuable.

Summary

With the strong implementation of Seedance 2.0 and Keling 3.0, the AI video track has officially moved beyond the 'novelty phase' of mere technical showmanship and entered the 'industrial realization phase' where commercial value explodes. The core insight of this technological revolution for the capital markets is that investment logic is shifting from upstream 'arms race in computing power' to deeper monetization of downstream 'IP asset exploitation'.

Investors should closely monitor companies that can be the first to utilize AI tools for tangible 'cost reduction and efficiency improvement' or successfully achieve 'cross-dimensional monetization' of IPs. As the barriers to content production are being shattered, those who possess non-replicable IP cores and build moats of copyright protection will emerge as the true winners in this wave of AI applications.

655K Views