Dialogue with Xinghuo Animation's Lv Shifeng: Content Remains the Core Variable in AI Comics

AI comic series’ first round of rapid development is no longer just a competition of technological efficiency. Over the past half year, improvements in model capabilities, platform traffic expansion, and integration with the commercial system for short dramas have together brought this new content category to the forefront. Low cost, short production cycle, and high payout have attracted a large number of teams to enter quickly, greatly amplifying industry supply in a short time. But more supply doesn’t mean more hits. Lü Shifeng, founder of Xinghuo Animation, recently stated in an exclusive interview with Wallstreetcn · All-weather Technology that if we define a hit as having over ten million views, the industry hit rate may still be about 4% in 2025, while now “even 1% is difficult.” In his view, AI technology has indeed lowered production thresholds, but it hasn’t simply led to continuously reduced costs. As model capabilities improve, market demands for content quality simultaneously rise. Xinghuo Animation’s growth path fits right into these industry changes. The company was founded in early 2024, initially focusing on text-to-comic tools and comic content supply. In the second half of 2025, as image and video model capabilities improved, the company shifted towards comic series production and distribution, expanding from a team of 20-30 to more than 200 members. Against the backdrop of rapidly expanding supply, Xinghuo chose the path of high-quality, serialized, and original IP development to resist the dilution of traffic caused by the influx of low-quality content. During the 2026 Chinese New Year season, Xinghuo released “Journey to the West: Mistakenly Treating the Jade Emperor as Dad,” which became a phenomenal hit: within 35 hours of launch, its Douyin views surpassed 100 million, with total views across all platforms reaching 230 million; it topped the Hongguo chart for 11 consecutive days and became the first comic series on the Hongguo platform to exceed a popularity score of 60 million. Centering on the “Journey to the West Universe,” Xinghuo has released six series so far, each with over ten million views. The new show released during the May Day holiday, “Journey to the West break-up group,” has had over 1 million favorites, and the upcoming third season of “Journey to the West Bricklayer” is close to 1 million reservations. This reflects the core change at the next stage of AI comic series: while traffic is amplified quickly, low-quality supply is being filtered out even faster. For production companies, merely pursuing low cost and high output no longer forms an advantage; the importance of high-quality series, serialized development, and ongoing distribution capability is rising. [image] Around the production model, platform changes, industry differentiation, and Xinghuo Animation’s next steps, All-weather Technology recently conducted an interview with founder Lü Shifeng. The following is an edited transcript. **Hits Are Not Mass-Produced** **All-weather Technology:** How did Xinghuo Animation originally enter the comic series industry? **Lü Shifeng:** Xinghuo was founded in early 2024, initially aiming to use AI technology to create generative comic content models and tools. At the time, text and video model capabilities were not mature enough. We decided to first focus on the comic field, using text-to-comic generation to solve production efficiency. So in 2024, we mainly focused on image models, investing in tech and computational power to develop our own model and tool products. By 2024 and the first half of 2025, our model tools’ efficiency and generation quality were relatively advanced in the industry. In the second half of 2025, as basic image model capabilities quickly improved, especially as general model generation became widely better, we started expanding our model and tool capabilities horizontally, moving from comic generation to comic series content generation. The company business also transformed. Previously, we mainly supplied comic content—at peak, 15,000 comics a month; by the second half of 2025, we started producing and distributing comic series content. Now, on one hand, we invest a lot of manpower in content production; on the other hand, we continue developing tool products and exploring some platform-based business. Although the delivered content has changed, the underlying logic is continuous. The workflow, content production pipeline, and tool capabilities accumulated in comic making are now transferred to comic series production. **All-weather Technology:** In the past half year, industry supply has grown fast—has your team expanded significantly? **Lü Shifeng:** Yes. We’ve grown from a team of 20-30 to over 200 members. **All-weather Technology:** With team expansion and technical progress, production capacity will rise fast. After increasing supply, how do you maintain the hit rate? **Lü Shifeng:** The core is still demanding high content quality. For content to become a hit, it starts with choosing the story, project approval, and topic planning. This industry develops quickly, so many content production teams came in. There are roughly two schools in the market: one produces large quantities at relatively low cost; the other focuses on high-quality content but with smaller supply scale. We clearly chose the latter. So we invest more in story planning, script quality control, and content presentation. Our overall hit rate is still good, about 30-40%. **All-weather Technology:** Will model capability improvements cause content production costs to drop faster? **Lü Shifeng:** Not much. As model capabilities grow, our requirements and standards for content quality also change. Input-wise, a project that used to need 10 people can now be driven by 5 or 6, but the timeline hasn’t changed that much. In terms of computational costs, we still aim to keep a work’s cost around 50,000 yuan to meet higher quality requirements. For major projects, we may not spare costs to produce. Only by supplying good content can users pay. **All-weather Technology:** There’s an idea in the industry that AI comic series are approaching “one person, one show per day.” Do you think this will come true? **Lü Shifeng:** Looking at future trends, “one-person companies” or “one-person studios” will definitely appear. But one show per day is a bit exaggerated. Producing a show requires a script, characters, scenes, props, special effects, and other information; this info needs to be converted into image production material, and then go through a video process. There’s still a process. These steps can eventually be done via Agents. But today, Agent-generated content is still some distance from high-quality, consumer-level content. So, one show per day is a bit exaggerated. Long-term, it’s more likely that a director can be solely responsible for developing a show. **All-weather Technology:** Do you think AI comic series can incubate native IPs? **Lü Shifeng:** Native IP is the direction we want to develop. In China’s anime series, perhaps 70-80% of works are novel adaptations. That’s because novels have a substantial existing audience, who partly overlap with anime series fans, which helps subsequent development. But today, with much more content supply, IP itself is a content selection logic—as creative output increases, chances to select IPs also rise. Content reaches more users, and through continuous incubation, content images or symbols gain wider user recognition. I believe comic series have the ability to incubate IP, linked to their product attributes—they have explosive traffic potential, where a work may exceed 100 million views in 30-50 hours, reaching a massive user base. If development continues, we can better operate these users. With continuous content updates, the vitality of the work grows stronger. We firmly believe comic series can incubate widely beloved market IP. **All-weather Technology:** How are Xinghuo Animation’s core “Journey to the West” series performing in terms of breakout success and long-tail effect? **Lü Shifeng:** “Journey to the West” is a national-level IP—audiences know the characters and relationships well. Our adaptation is a group story; for users, it’s a humorous plot involving familiar character relationships, making understanding and viewer immersion higher. Comedy fans are a big group now—data shows 40% female viewers, so it’s quite a breakout. **All-weather Technology:** What is Xinghuo’s strategy for overseas release of comic series? **Lü Shifeng:** We’ve already released some comic series overseas, mainly in the US and Japan. User recognition, payment, and repeat purchase rates are all quite good. That relates to the market itself. Animation production capacity was limited, but anime and animation markets have huge user bases with annual consumption over $10 billion. Many over-$10-billion tracks are red oceans, but in animation I think it’s still a blue ocean—content supply is limited and the user scale is large. So, when supply increases, the market ignites quickly. From content supply to consumer demand to closing commercial chains, the growth flywheel turns very fast. **Clearing Out Will Eventually Arrive** **All-weather Technology:** After supply increases, what changes are there in the industry hit rate? **Lü Shifeng:** There’s more industry content, but the industry data hasn’t improved. In 2025, with Douyin views over ten million as a hit standard, the hit rate was about 4%. Today, even 1% is difficult. This trend may not change much this year. For us, our content and distribution hit rate is still around 30%. Our core response is sticking to high quality content. I think users will still pay for quality. **All-weather Technology:** The commercialization system for comic series is largely integrated into the short drama system, especially ByteDance platforms. What’s your view? **Lü Shifeng:** Any category ultimately depends on the scale and reality of user demand. Domestically, ByteDance’s support for the industry and efficiency of moving from traffic to commercialization does relate to industry growth. But the same content is pay-per-view overseas—users must pay. From the payment model’s perspective, the data is not worse than domestic. **All-weather Technology:** What are the different platform strategies for comic series? **Lü Shifeng:** The positioning differs. Tencent has accumulated many anime series users; ByteDance platforms like Douyin or Hongguo, have the largest traffic. For content companies, we must consider what content fits which platform. From a business perspective, we have to judge which content performs best on which platform. So we work differently with each. With Tencent, we bring more anime-style or 3D content; but now, our greatest collaboration is still with Douyin, and most content goes there. **All-weather Technology:** Is having a platform background decisive in getting traffic or IP? **Lü Shifeng:** Not necessarily. In content industries, users follow content. Platforms always need good content. So partnerships are mutual, not simply relying on a platform. If you can’t produce good content, the platform can easily drop you. **All-weather Technology:** Do you chase trending topics or proactively create them? **Lü Shifeng:** Predicting themes is very hard. One strategy is to spot a hot theme, quickly make something similar—even producing a series in 5 days. This may catch the market, but if you keep doing it, you’ll exhaust your creative capability and learn nothing about content creation. So, we build our own IP matrix in mainstream categories to balance market traffic fluctuation; meanwhile, we stick to high-quality content to win over users with excellence. We’re now discussing IP development—perhaps first building a big world-view, a foundational structure. Once the basic project foundation is built, we can develop story branches and produce more works. Having such a system lowers production risks. **All-weather Technology:** Why is this industry’s iteration so fast? **Lü Shifeng:** Because technology lowers entry barriers. AI comic series develop quickly thanks both to user demand and the transformative power of modeling, causing loads of newcomers. Even a 5-person team can make a series, and it could bring 20 or 30 times the returns. Algorithms have some effect, but the core is modeling capability fueling industry change and enabling many content creators to enter. **All-weather Technology:** Where might the next key market change come from? **Lü Shifeng:** When content scale gets big enough, lots of low-quality content can’t earn money and will be cleared out, and the market will be restructured. The industry’s logic follows people, people follow money. When they can’t earn, they’ll leave the industry. Then, more high-quality content will compete. This year should see obvious differentiation. There will still be some small players, but the market will focus on leading companies. Top companies are better at content creation, story structure, pacing, technical buildup, and traffic investment. The future industry may see an 80/20 split, with 80% of returns going to comprehensive output companies. **All-weather Technology:** What’s Xinghuo’s business scale plan this year? **Lü Shifeng:** Distribution business this year will reach about 300–400 million yuan. **All-weather Technology:** Will the company continue to seek financing and expansion? **Lü Shifeng:** We have already completed two rounds of financing, totaling over $10 million, with some capital reserve. Looking long-term, we will continue financing, with domestic and overseas distribution progressing in parallel. Risk Warning and Disclaimer The market entails risk, and investment must be cautious. This article does not constitute personal investment advice, nor does it take into account the special investment objectives, financial situation, or needs of individual users. Users should consider whether the opinions, views, or conclusions in this article meet their specific circumstances. If investing based on this article, responsibility is borne personally.