The 'one-person company' trend is exploding: some earn millions annually, while others see incomes shrink by 90%

Source| Tech Planet

Text| Zhai Yuanyuan

The 'one-person company'—a model in the AI era where individuals independently launch businesses using digital tools (OPC)—has gone viral.

2026 is being hailed as the 'Year One of the One-Person Company.' Empowered by AI, ordinary individuals are entering a golden age of entrepreneurship. With just one person plus AI, it’s now possible to start a business at minimal cost—using large language models to write code, GPT to generate content, and AI to monetize. AI is no longer just a technological concept; it’s now genuinely creating commercial value for individuals.

Wealth creation stories around 'one-person companies' are emerging one after another. Dozens of cities across the country have rolled out policies encouraging and supporting entrepreneurship through one-person companies. According to publicly available data, by mid-2025, the number of active one-person companies nationwide had exceeded 16 million, accounting for 27.4% of all enterprises in China.

Of course, this boom has also sparked considerable controversy. Some reports claim that 'the first wave of one-person company founders has already exited.' Critics argue that there is no fundamental difference between 'one-person companies' and earlier entrepreneurial models like 'sole proprietorships' or 'curator-led ventures.' Statistics even suggest that behind the OPC (One-Person Company) frenzy, the reality is that 52.7% of these entrepreneurs earn less than 7,000 RMB per month.

In fact, entrepreneurship always carries risk, and only a minority ever succeed. AI has lowered entry barriers and improved cost-efficiency, but it does not guarantee success. What is the real survival situation for one-person companies, and who has actually captured the benefits? We spoke with several entrepreneurs to find out.

Developed six bullet-hell games, earning one million annually

Entrepreneur: Mr. Zhang, game development sector

Since leaving my job to start my own business in March 2025, I’ve developed six bullet-hell games over the past year, generating nearly 20 million RMB in total gross revenue, with personal net income of approximately 800,000 to 1 million RMB. My most successful title alone has brought in over 8 million RMB in cumulative revenue.

Platforms like Douyin take a high commission—typically 50%—with streamers receiving 42%, leaving developers with only about 8% of the gross revenue.

The gaming industry has long suffered from 'high costs and long development cycles,' problems perfectly addressed by the one-person company model. The production cost for a single game can be kept between 1,000 and 1,500 RMB, primarily covering AI tool subscriptions and asset pack purchases. Compared to traditional teams that often spend upwards of 2 million RMB on development, this represents extreme cost compression.

Unlike traditional mobile or Steam games, bullet-hell interactive games leverage built-in platform traffic, requiring virtually no additional marketing or promotional spending—a zero-marketing advantage that significantly lowers the barrier to entry for entrepreneurs.

AI tools have replaced traditional art and programming teams, enabling highly efficient production. During game development, AI handled roughly 70% of the artwork for me. I’ve established a complete workflow: 'Doubao (keywords) → Banana (concept art) → Triple3D/Paid 3D (modeling) → Blender (mesh optimization) → Mixamo (rigging).'

AI has also enabled code-assisted development, allowing AI to write Java code with an accuracy rate of up to 99%, significantly lowering the barrier to software development. Aside from AI, I personally focus primarily on final integration, UI adjustments, and engine compatibility work.

In terms of game development timelines, AI has effectively compressed production cycles. Traditionally, a team of around 10 people would require two to four months to develop a similar game, whereas using an AI-powered toolchain, a solo developer can complete the project in approximately 15 days. Moreover, the final product quality is nearly on par with that of major studios—indistinguishable from outputs by traditional game development teams—and ordinary players or streamers can hardly tell whether it was AI-generated.

Of course, solo game development entrepreneurship demands certain core competencies from the founder: the one-person company model suits professionals who either 'have distribution channels (client resources)' or 'understand workflows and possess strong creative ideas.' AI cannot replace creativity or project management skills. Experience remains the strongest entrepreneurial moat—while AI lowers technical barriers, individuals with zero industry experience struggle to manage the process effectively, as they lack the ability to assess the correctness of AI-generated content and make necessary corrections.

The gaming industry is currently experiencing severe layoffs, making it difficult to secure traditional roles. However, the adoption of AI tools has empowered individuals with productivity levels comparable to entire teams, creating a window of opportunity for experienced professionals to pursue independent entrepreneurship.

For me personally, in addition to developing my own games, I also take on outsourced projects using a mature AI-driven production pipeline. Each project is priced between RMB 30,000 and 50,000, with a delivery timeline of about one week, primarily to ensure stable, reliable cash flow.

My work relies on AI constantly, and my income is three times that of an average white-collar worker my age.

Entrepreneur: October, materials R&D sector

I am an entrepreneur running a one-person company in Japan and hold Japanese permanent residency. Previously, I worked in the traditional materials research and development industry. I founded my solo venture in July last year, operating in deep processing plus retail (customized services), and the business has been running for less than a year.

My current monthly income is roughly three times the average for Japanese white-collar workers of my age, with annual earnings of approximately 15 million yen—significantly more than what I earned in my previous salaried role.

Regarding my motivation to start a business, I decided to launch my venture only after witnessing the emergence of generative AI. With AI augmentation, ordinary individuals now possess the capability to operate independently—90% of issues can be resolved simply by querying AI tools. Today, my daily workflow is almost entirely dependent on AI tools for tasks such as customer service translation, messaging, image generation, order processing, programming, and legal consultation.

Currently, my monthly costs for using AI tools are relatively low—approximately 3,400 JPY per month. My work does not involve expensive API calls, and my main use cases are as follows:

1. Language localization: I use ChatGPT to proofread, refine, and localize translations involving specialized terminology, addressing linguistic differences between Chinese and Japanese.

2. Knowledge acquisition and learning: When encountering unfamiliar topics in entrepreneurship—such as taxation, regulations, manufacturing know-how, or business judgment—I leverage AI to quickly research and synthesize information to support decision-making.

Compared with the trend of 'one-person companies' in China, such solo entrepreneurship is less common in Japan—but it has been growing steadily, with the number of Japanese individuals running one-person companies increasing exponentially. For foreign nationals, however, the barriers to starting a business in Japan remain relatively high: applicants must obtain a Business Manager visa, which requires an initial investment of at least 30 million JPY. Without permanent residency or a stable work visa, the entry barrier becomes extremely steep. Industry-specific licensing requirements also vary widely across sectors; for example, food service, medical aesthetics, and secondhand goods trading all require specific operating permits. Fortunately, the industry I operate in has comparatively relaxed regulatory requirements.

Japan’s industry barriers are notably high. In the Japanese market, establishing personal or corporate creditworthiness is a critical prerequisite for accessing supplier networks. The market is highly insular—breaking into any given industry typically requires an internal referral from an established 'gatekeeper.' Without such an introduction, building trust is nearly impossible, and gaining access to suppliers involves rigorous credit vetting procedures. In my early days as an entrepreneur, business progressed slowly; however, over the past six months, word-of-mouth referrals have driven exponential growth, and my operations have now stabilized.

Compared with China, Japan’s consumer market adopts new technologies and innovations—such as smart tools or digital solutions—more slowly. Even QR-code-based ordering only became popular in recent years. This lag creates opportunities for Chinese entrepreneurs to fill market gaps. Competitive dynamics also differ: Japan’s market is less prone to the intense 'involution' seen domestically. With a smaller population and generally milder competition within industries, one-person companies can carve out viable niches. Low-price strategies tend to fail in Japan, as consumers often view cheap products with skepticism; conversely, premium pricing can more effectively build trust.

AI handles 80% of repetitive tasks, doubling my revenue

Entrepreneur: Xiao Tao, biotechnology sector

I launched my venture in August last year. As a one-person company, I didn’t initially seek out clients—instead, I was introduced by former colleagues to customers in the biotechnology field. I secured clients first, then decided to formally establish a one-person company to fulfill their needs.

The industry has notable entry barriers. I primarily provide B2B services focused on biological cell culture and development, with a core competency in using algorithms to optimize cell culture media formulations and enhance predictive efficiency. My prior experience at a biotech firm equipped me with AI algorithm expertise; that company specialized in culture media development—a field pursued by only a handful of institutions globally—leveraging AI to predict optimal formulations, thereby creating a significant technological moat.

Compared to a full-time job, my solo-founder business model has doubled my income (approximately two-fold) while significantly increasing my time flexibility, resulting in a much higher return on effort.

My core daily workflow relies on the DeepSeek model, combined with the Claude model to build AI agents. I use VS Code as my development environment and automate tasks through these agents.

Computing costs are actually quite low—my current spending on DeepSeek tokens is less than 30 RMB, and my only hardware investment is my personal computer, making the overall startup cost extremely low. AI agents handle roughly 80% of routine, repetitive tasks (such as data analysis and report generation), while the remaining 20%—core judgment, framework design, and client communication—is primarily managed by me.

Twenty percent of the work cannot be replaced by AI. Although AI handles a large portion of the workload, entrepreneurs must continuously monitor the agent’s progress and output direction to prevent deviation from goals and ensure delivery quality.

In my view, roles such as data analysts, outsourced product managers, and administrative staff—whose work heavily relies on digitized, repetitive tasks—are at significant risk of direct replacement by AI agents. At the same time, these professionals are also well-suited for entrepreneurship: those with technical backgrounds, business acumen, and client networks are ideal candidates to transition into solo-founder businesses. If they master AI tools and already have client relationships, they should consider launching their own one-person company.

I’ve been running my business for less than a year and have encountered numerous challenges, exposing certain operational gaps. Founders need to address non-technical weaknesses—such as interpreting government policies, navigating company registration procedures, and ensuring tax and financial compliance—typically by leveraging external resources. I advise against blindly jumping into solo-founder entrepreneurship; introverted individuals face inherent disadvantages in customer acquisition and fundraising pitches and must secure stable client relationships before starting a business.

Income has dropped by 90% compared to working at a big tech firm—do not follow the trend into entrepreneurship without careful consideration.

Founder: A Yuan’s Free Voyage, Cross-border E-commerce

I previously worked in operations at a major internet company. With industry growth slowing and an invisible career ceiling looming, I decided to leave and start my own business focused on cross-border e-commerce targeting Latin America.

Due to prior work assignments, I was posted to Latin America and became quite familiar with the market. Given the rapidly growing trade ties between our regions and the market’s immense potential, I chose to focus on Latin America. Because of limited startup experience, I initially joined an existing startup team. In April this year, I officially went independent, adopting a lightweight solo-founder model and managing the entire process—from product selection to after-sales service—on my own.

In my personal case, the capital threshold to start a one-person company isn't high: company registration is free, accounting services cost around 2,000 RMB per year, e-commerce platform onboarding is generally free, and depending on the chosen business model, you may need to purchase a local Mexican tax ID for under 10,000 RMB.

Currently, I’m primarily targeting the Mexican market using the 'overseas warehouse' model—pre-stocking inventory in the platform’s official fulfillment centers to reduce logistics costs and improve delivery efficiency.

AI has replaced approximately 60% of manual workload. For lean cross-border e-commerce startups, product selection and product page optimization are the most critical tasks—and nearly all of this can now be handled by AI. AI tools cover market research and product selection, generation of text/image/video content, optimization of product detail pages, and financial data analysis.

In my personal workflow, I mainly use GPT, Gemini, and Doubao’s video generation tools, with an average monthly subscription cost of just over 200 RMB—significantly reducing operational expenses compared to traditional human designers. Typically, professional product image design alone would cost as much as 500 RMB per item.

Of course, I don’t recommend quitting your job impulsively to start a business. There’s excessive hype online about 'one-person companies,' and you shouldn’t follow the trend blindly. Working at a major tech firm still offers excellent value. Previously, I earned an annual salary of around 600,000 RMB at a big tech company; now, after two months of entrepreneurship, my sales total just over 10,000 RMB, with a profit of about 1,000 RMB. Even including income from freelance store management, my earnings have dropped drastically—by roughly 90%.

That said, the growth value gained from entrepreneurship truly surpasses several years of salaried work. As the Western saying goes: 'Get your hands dirty.' Starting a business has forced me to deeply understand operations firsthand—shifting from a platform-side perspective to on-the-ground execution. I’ve gained a visceral appreciation for the hardships of entrepreneurship and developed a more grounded understanding of business fundamentals, which has made me far more down-to-earth as a person.

728 Views