Original · Unique Research · 2026-01-22
Historical edition: This complete textual article and all 12 selected interview Q&As were published on January 22, 2026. Production capacity, costs, views, business growth and market assessments are the original author's and Wu Wanxuan's statements at that time, not independently verified current performance. “AI photorealistic dramas” translates the source's AI真人剧: AI-generated productions resembling live-action, not footage necessarily filmed with actors. The source compares total traditional-production budgets with AI costs per minute; these are not like-for-like cost units. Its 40 minutes in 10 days equals 4 minutes a day, while its quoted 4–5-minute range is retained. Chinese organization and brand names are transliterated, not claimed as official English names. Images and image-only material are outside this text-only edition.
While Everyone Bets on “the Next Douyin,” This Company with 13 Years in Overseas Markets Is Quietly “Selling Shovels”
—Xiaoying Technology's Business Playbook for AI Short Dramas: Do Not Build Models; Build Scenarios
If you still think taking short dramas global means translating a domestic hit, dubbing it, uploading it to YouTube, and collecting revenue, you may not even have found the door.
In the overwhelming AI wave, those most anxious are not empty-handed entrepreneurs but veterans carrying heavy burdens. Transforming means courting death; refusing to transform means waiting for death—this seems to have become the industry consensus.
What AI short dramas truly change is not cost, but who is qualified to tell stories.
01. The Opportunity for Small and Midsize Teams: The Greatest Danger Is Not Competition, but Thinking the Opportunity Is Upstream
When many people hear “AI sector,” their first reaction is: “Should I train a foundation model?” Wu Wanxuan rejects the idea outright.
For small and midsize teams, foundation models are a game for giants; the real opportunity lies at the application layer. Yet that does not mean building another chatbot. The valuable part is often not the foundation itself, but who can turn foundational capabilities into a workflow that can deliver, be reused, and generate revenue.
Wu Wanxuan identifies two key directions:
• AI-native content generation—music, images, and video—a path already mature in audio and images and now extending into video
• AI assistants plus vertical fields, including scenario-based applications such as healthcare and emotional stress relief
Models are like electricity; applications are like lamps. Electricity certainly matters, but first let people see the light. You do not need to understand the underlying Transformer architecture. You need to understand how to use AI to generate a healing video late at night for someone whose heart has just been broken.
That is the dividend for ordinary people: packaging complex technology into a foolproof service.
02. Three to Five People Make Fantasy Dramas: Redefining the Combat Power of a “Small Team”
Over the past year, I have heard too many grand narratives about “AI replacing film and television professionals.” With Wu Wanxuan, however, I saw a series of small, concrete arithmetic problems.
What is the logic of traditional film and television production? It is a capital-intensive game. To film a respectable urban drama, you need a crew of dozens, a month of work, and costs in the millions of yuan. For a small or midsize team, that is an insurmountable wall.
On Xiaoying's AI production line, however, the logic has been completely rewritten:
a team of 3 to 5 people can produce a polished 40-minute photorealistic AI drama in 10 days. That equals 4-5 minutes per day. The cost? Several thousand yuan per minute.
Notice that this means a finished work, not a rough Demo. With three such premium production teams, the production line can truly accelerate.
What does this mean? A geometric release of production capacity. Previously, teams competed on capital and resources; now they compete on imagination and storytelling. Costs fall directly from millions of yuan to thousands of yuan, a dimensionality-reduction strike that places the admission ticket to content creation in ordinary creators' hands.
This is not merely an increase in efficiency, but a devolution of power.
03. The Real Battlefield for AI Photorealistic Dramas: Film Stories That Were Previously Unaffordable
If AI merely makes filming cheaper, it is at most a low-cost substitute. Wu Wanxuan's insight goes much further.
She states the secret directly: the best battlefield for AI photorealistic dramas is not head-on competition with traditional productions over family ethics or wealthy-family revenge, but territory where live-action filming is unaffordable.
Fantasy, time travel, rule-based horror, zombies, post-apocalyptic wastelands, and cyberpunk—subjects filled with visual spectacle and requiring expensive special effects—were previously special-effects furnaces that only top studios dared approach. If small teams tried, their productions were likely to become jokes about cheap effects.
Today, however, these subjects occupy AI's comfort zone. Have live actors film a city besieged by zombies? Special-effects spending can burn through the floor. AI excels at precisely this.
AI is therefore not a budget replacement for live-action filming, but its special forces, sent specifically to attack the hardest subjects. It does not imitate reality; it expands reality's boundaries, making it possible to realize at low cost wild ideas that could once exist only on paper.
04. The Truth About Going Global: The Devil Is in the Subtitles
When the subject turns to globalization, this is Xiaoying's home territory after 13 years of experience.
A fatal misconception is widespread: many small teams think taking short dramas global is simple—when a drama succeeds domestically, quickly buy the rights, translate the subtitles, dub it, upload it to YouTube, and wait for dollars to arrive.
Wu Wanxuan not only pours cold water on that idea, but offers a guide to avoiding traps drawn from painful lessons:
First barrier: the copyright trap
Domestic distribution rights ≠ overseas distribution rights. Get this step wrong and everything afterward becomes litigation. Many teams fail at the first step without even determining whether they are entitled to distribute the content abroad.
Second barrier: the localization chain must be stable
Translation, dubbing, and encoding must remain controllable throughout the entire process. One exceptionally consequential detail is subtitle processing.
Wu Wanxuan emphasizes: “Many teams crudely place foreign-language text over Chinese subtitles. This patchwork approach directly damages immersion for overseas viewers. You must erase the original subtitles seamlessly before encoding the new ones.”
The detail sounds trivial, but it exposes the difference between professionals and hobbyists. Cleaner subtitles bring more revenue; this is the invisible moat of global short dramas.
Third barrier: do not target the whole world immediately
For small teams with limited budgets, Wu Wanxuan offers extremely practical advice: do not imitate major companies by trying to reach every market and every user immediately. That is not ambition; it is running yourself to death.
Hold tightly to one specific category and one specific country:
• Southeast Asia: rural supernatural stories, rule-based horror, and large-scale effects
• Japan and South Korea: romantic shots and sweet romance
• Western countries with predominantly white populations: werewolves, vampires, and localized motifs
The work does not end after translating a Chinese drama into English. You are translating the routines of one content industry into the emotional buttons of another culture.
Likewise, the urge to build an independent website or APP immediately may sound sophisticated, but that is a game for large amounts of capital. Borrow a boat before going global: first make the process work on social media through YouTube Shorts/TikTok channels. That is the route to survival.
05. Tools and Content: One Is the Foundation, the Other the Ceiling
What intrigued me most in the interview was the transformation of Xiaoying's identity.
The greatest change was moving from consumer products to business services and establishing Yinghuo Engine. Yinghuo initially focused on technical solutions, then created its own studio and began producing content directly.
This is what Wu Wanxuan calls “eating your own dog food.” You cannot stand onshore teaching others to swim; you must jump into the water and swallow a couple of mouthfuls yourself.
By producing its own content—AI animated comics and AI photorealistic dramas—Xiaoying is effectively conducting intensive combat exercises. It refines tools while validating the business model. Reported fivefold growth from 2023 to 2025 proves that this dual-engine path of tools plus content works.
Asked whether tools or content revenue sharing make a better business, Wu Wanxuan answers almost without hesitation: the latter will inevitably win in the future, but tools form the foundation.
Behind that statement lies a more mature commercial rhythm:
• At the tool layer: API integrations, including direct integrations by many major companies; workbenches for social-media teams; and a delivery model for special case scenarios, combining AI output with human review and optimization
• At the content layer: productions made in partnership with platforms, combining guaranteed production fees with revenue sharing after release
Tools are like a harbor; content is like the open ocean. A harbor may not be the most glamorous, but it keeps ships entering and leaving. The open ocean is more profitable, but you must truly know how to sail and withstand storms.
This offers a warning to every AI entrepreneur: the era of selling pure SaaS tools may be ending. Future giants will understand both how to make the weapons and how to fight the war.
06. The Verdict Three Years from Now: Will AI Content Move “from Controversy to Habit” Like E-Books?
At the end of the conversation, Wu Wanxuan offered a moving analogy.
She said that when e-books first appeared, traditional readers dismissed them, believing only words printed on paper qualified as books and that text on a screen lacked a soul. Today, no one agonizes over the medium; we care only whether the words move us.
AI short dramas will be the same.
Three years from now, when we revisit today's debate over AI generation versus live-action filming, we may find it amusing. Audiences will not care whether an image was filmed by a camera or calculated by a graphics card. They will care only about this:
during this several-minute story, did you make me cry, laugh, or feel my heart race?
What excites Wu Wanxuan most is not the technology itself, but that small and midsize teams and individual creators finally have an opportunity to produce content and can move from written creation toward commercialization. Xiaoying wants not only to serve enterprises, but to provide foundational services to individual creators pursuing commercial goals.
Behind this lies a grander narrative: when AI reduces costs and one person can support an entire production crew, the power to create content returns from capital to creators.
That will be the real carnival.
Closing Thoughts
• AI's reduction of costs is only the surface; what it truly lowers is the barrier facing creators.
• Tools determine whether you can enter; content determines whether you can remain and share the rewards.
• Technology will eventually fade into the dust; only good stories will mingle with the light.
About Xiaoying Technology: a video-editing technology company with 13 years of global operations that provides various AI technology solutions. It has mature deployments in AI animated comics and AI photorealistic dramas and achieved reported fivefold growth from 2023-2025. Its film-and-television content company has mature production processes and teams and is advancing an AI-native drama production workbench for global markets.
Selected Interview Q&A
Q1: What does Xiaoying Technology do?
Wu Wanxuan: Xiaoying Technology is a technology company with 13 years of global operations. During its first ten years, its main business was content-creation apps. In 2024, it established the business-facing brand Yinghuo Engine, which includes the AI short-drama business, and created Yinghuo Studio with mature film-and-television production processes and teams.
Q2: What revenue opportunities remain in the AI sector for small and midsize teams?
Wu Wanxuan: For small and midsize teams, the opportunity lies in the AI application layer, primarily in two directions:
the first is AI-native content generation, including music, images, and video. Generation applications for music and images are already mature; the future will focus on video generation and creation, which will be a highly promising field.
The second involves AI assistants, or AI Assistant products, combined through AI with specialized fields. These include serious areas such as law, finance, and taxation; more entertaining areas such as AI healthcare; and applications related to stress relief. They are AI fields where we already observe activity and expect substantial future development.
Q3: Can you share production-capacity data for Xiaoying's AI short-drama business?
Wu Wanxuan:
For AI animated comics:
we make neither motion comics nor crude parody animation; we produce only premium 2D and 3D animated comics. By a very conservative estimate, we can stably produce 10 premium 2D/3D animated-comic series per month.
For AI photorealistic dramas:
We tested on Douyin and achieved more than one million organic views in under 2 days. At our current capacity, one team of 3-5 people can produce 40 minutes in 10 days, equivalent to 4-5 minutes per day. We currently have three premium production teams capable of scaling the work.
Q4: What advantages do AI photorealistic dramas have over traditional live-action filming?
Wu Wanxuan: There are three main areas:
Cost advantage: AI photorealistic dramas can cost only several thousand yuan per minute, a major reduction compared with the million-yuan cost of premium live-action production.
Efficiency advantage: a team of 3-5 people can produce 4-5 minutes per day, an enormous efficiency gain over live filming.
Subject-matter breakthrough: most importantly, AI photorealistic drama does not disrupt live filming primarily in family or revenge stories, but in subjects difficult to film with live actors—fantasy, time travel, rule-based horror, zombies, and suspense. These are difficult for live-action production but are precisely where AI photorealistic drama has an advantage.
Q5: How do domestic and overseas AI short dramas differ in subject matter?
Wu Wanxuan: It depends on the overseas region:
Southeast Asia: choose subjects similar to those suited to China, such as rural supernatural stories and rule-based horror, which work well in Southeast Asia.
Western countries with predominantly white populations: strongly align with local material, including werewolf and vampire stories, which may require greater attention.
The core difference is that for overseas subjects, AI photorealistic drama is used more to reduce the filming cost of local productions dramatically. The logic differs from China, where the focus is more on filming subjects that are otherwise difficult to realize.
Q6: Which traps are small teams most likely to encounter when taking short dramas global?
Wu Wanxuan: Small teams must avoid several misconceptions:
First, copyright: holding domestic distribution rights does not mean you hold overseas distribution rights. This is the most important issue. Many people fail to handle the first step and encounter many problems later.
Second, the localization chain must be stable: you need reliable delivery across the complete process of translation, dubbing, and encoding. Xiaoying has continually developed this at the tool layer.
Third, do not target the world immediately: you cannot imitate major companies by targeting every market and every user from the start; doing so only runs your company to death. Hold tightly to one specific category and one specific country.
Fourth, channel selection: for small teams, social media such as TikTok/YouTube is the lowest-cost channel. Do not immediately think about building owned audience channels or an APP; costs expand without limit in that situation.
Q7: Which details are easily overlooked during production of short dramas for overseas markets?
Wu Wanxuan:
Cultural taboos matter greatly: when entering the Middle East, for example, pay close attention to local culture and taboos.
Subtitle-processing details: many teams leave Chinese subtitles underneath and simply add another layer of translated subtitles above them. For audiences in some regions, that severely harms the viewing experience. Seamlessly erasing the original subtitles before encoding new ones is a small but important detail.
Dubbing strategies must vary by scenario: a hit drama may need highly refined AI multi-character dubbing with strong emotional expression. If a channel is already mature and has lower requirements for dubbing quality, however, it needs the ability to localize large numbers of dramas rapidly and release them quickly.
Q8: What is Xiaoying's business model?
Wu Wanxuan: We operate at two levels:
Tool layer:
API integration: many major companies directly integrate our API services
Workbench: used by social-media teams publishing on YouTube or TikTok
Delivery model: for special case projects, after AI tools produce the output, we apply a stage of human review and optimization
Content layer:
We partner with platforms on productions and receive guaranteed revenue sharing—first, a guaranteed production fee, and second, a share of revenue after the platform release. This is our primary source of revenue.
Q9: Which makes a better business: tools or content revenue sharing?
Wu Wanxuan: Looking to the future, it will inevitably be the latter—content revenue sharing—but tools form a foundation and offer a good way to understand the current short-drama market ecosystem.
Q10: How did Xiaoying grow from 2023-2025, and why?
Wu Wanxuan: We essentially achieved fivefold growth.
This year has seen explosive growth in translated and dubbed dramas. Besides the rising number of participants entering the global short-drama business, another important reason is the rise of short dramas on YouTube. This has enabled many domestic short-drama producers that previously lacked the ability to build their own APP to enter overseas markets, causing localization demand to increase dramatically.
Q11: What is the greatest change after AI enters short-drama production?
Wu Wanxuan: I believe there are three greatest changes:
First, higher production efficiency: AI allows a team of 3-5 people, or even one person, to produce an entire drama independently. It does not simply replace traditional teams; traditional teams can also use AI to improve filming efficiency, including in special-effects scenes and CG work.
Second, more diverse subject matter: live-action filming of subjects such as rule-based horror and post-apocalyptic settings either costs enormous sums or cannot guarantee quality. AI makes real diversity in subject matter possible.
Third, greater advertising-material capacity: it provides short dramas with more creative assets for global distribution and paid acquisition.
The first point excites me most—small and midsize teams now have an opportunity to produce content. Individual creators can move from written creation toward commercialization. We want to serve not only enterprises but also individual creators pursuing commercial goals, providing foundational services for the broader environment.
Q12: How will people's assessment of AI short dramas differ three years from now?
Wu Wanxuan: I think it resembles several years ago, when people believed reading meant using paper books, that only books printed on paper were good books, and that e-books were inappropriate for a proper reader.
Years from now, people will regard content made with AI as high-quality content as well, rather than assuming AI-made work lacks qualities of live-action production. I hope we reach a state where people focus more on the plot itself instead of agonizing over which technology produced it.