On February 11, following the viral success of Seedance 2.0, related concept stocks experienced a divergent market performance. C3.ai, Inc., Huanrui Century, and Decai Shares hit their third consecutive limit-up, while Enlight Media rose over 7%. In contrast, the AI training data, film and television, and AI short drama sectors collectively suffered heavy losses. At the time of writing, Xinhua News Agency fell more than 9%, Alpha Group, Chinese Online, and Reader Media dropped over 6%, with Visual China Group and China Publishing also declining over 4%. Hengdian Film and Jin Yi Film hit the跌停 limit-down, while Wanda Film and Happiness Blue Ocean fell more than 8%.
On February 10, Hengdian Film issued an announcement stating that its current stock price has severely deviated from the company's fundamentals and faces the risk of a rapid decline at any time. Should the stock price experience further abnormal increases, the company may apply for a trading halt for review. The company clarified that its AI short drama business is currently still in the investment and production phase, with both market acceptance and profitability models remaining uncertain, and the business has not yet generated any related revenue.
In early February 2026, the quiet internal testing of ByteDance's AI video generation model, Seedance 2.0, sent shockwaves through the technology and venture capital circles. Feng Ji, founder of Game Science and producer of "Black Myth: Wukong," hailed it as "the strongest video generation model on the planet, bar none," and asserted that it marks "the end of AIGC's childhood era."
The capital market reacted swiftly to this technological breakthrough. On February 9, the A-share media sector surged, with data showing the cultural media sub-index rising 4.79% that day, leading the market. Stocks like Chinese Online, C3.ai, Inc., and Jiecheng Shares achieved 20% or 10% limit-up gains. The rally continued on February 10, with over a dozen stocks, including Doker Culture, Rongxin Culture, and Visual China Group, collectively hitting limit-up.
Reportedly, Seedance 2.0 represents a comprehensive iteration over version 1.0. It employs a dual-branch diffusion transformer architecture capable of simultaneously generating video and audio. Using just a single image or a detailed text prompt, it can produce a multi-shot sequence video with native audio within 60 seconds. Its unique multi-shot narrative function automatically maintains consistency in characters, visual style, and atmosphere during scene transitions without manual editing, making it suitable for various scenarios like smart long-form video, film and television, e-commerce advertising, and domestic animation games. Version 1.0 was primarily suited for short video creation.
However, alongside the impressive technology, practical testing sparked controversy. In a test video published by TIM (Pan Tianhong), founder of the popular channel "Film Storm," he praised the model's detail, smooth camera work, and audio-visual synchronization, calling it an "AI that will change the video industry." Yet, he repeatedly expressed concerns—noting that without providing any prompts or audio files, merely uploading a personal facial photo allowed the AI to accurately replicate his voice; uploading just a frontal photo of a building resulted in the AI generating a rear view almost identical to reality. Throughout the test, TIM used the word "terrifying" six times to voice his concerns about the technology's boundaries, explicitly stating he had not authorized the use of his personal materials to the official team nor received any payment.
Following the controversy, on February 9, the Dream Operations team announced that, to ensure a healthy and sustainable creative environment, Seedance 2.0 would urgently suspend the function of using real-person materials as primary references and was optimizing the model based on user feedback.
Despite the controversy, the technological breakthrough of Seedance 2.0 is widely viewed positively by securities firms, which believe it will accelerate the development of the AI multimodal industry chain. Huaxin Securities pointed out that the model pushes AIGC into a Kardashev-scale leap in content production, potentially transitioning content creation from a manual era to an automated one, achieving a shift from traditional Type I production models to AI-driven Type II models. Guosheng Securities described it as a "game-over" level product, estimating that under neutral assumptions, it could reduce the "card-drawing" frequency in AI video generation by 50% and lower the cost per second of generation by 37% compared to peers, with a conservative estimate still showing a 5% reduction, effectively addressing cost pain points in the animation and drama industry. Bank of China Securities stated that the qualitative leap in Seedance 2.0's capabilities pushes video generation from simply "generating a sequence of frames" to "completing a work." Its four core abilities, including self-storyboarding and self-camera work, can provide average users with a near-"director-level" creative experience while significantly lowering the production threshold and costs for AI-generated animations and dramas, promoting industry-scale development. Additionally, the high computational demands of multimodal generation are expected to benefit upstream hardware infrastructure simultaneously. A research report from Founder Securities analyzed that the release of Seedance 2.0 confirms the AI application industry cycle entered an upward景气 trend in 2026, with major tech companies' technological efforts serving as a core catalyst, benefiting sub-sectors like AI-generated animation/drama, film and TV IP, and data elements.