Original · Unique Research · 2025-11-03
Editor’s note: This complete English edition preserves the original report’s commentary and the interviewee’s claims as of November 3, 2025; model coverage, pricing, customer counts and service-performance statements are not independently audited results. Jingling is Li Jinglin’s nickname. Blue Lion, Xinshu Network and Deci Mobile are translations or romanizations of the Chinese names in the source, not independently verified official English names. The source spells one music model “Uido”; that spelling is retained rather than silently corrected. Its “99% SLA” wording supplies no measurement period or further service-level definition. No publication location is inferred.
In today’s dazzling, fast-changing AI era, it can seem as though every clear-eyed entrepreneur is wrestling with the same dilemma: on one side stands a technological utopia that can be seen but not touched; on the other, the ever-scorching reality of survival beneath their feet.
Li Jinglin, founder of DeerAPI and widely known by the nickname “Jingling,” clearly belongs to the latter camp. He does not talk about lofty technological visions floating in the clouds. Instead, he is more like a blacksmith who forges AI tools into instruments of survival, drawing on hard-earned experience to explain how to stay alive before the tailwind reaches you.
Li Jinglin’s story is by no means that of an “AI-native” founder. He was among China’s earliest independent webmasters and has extensive entrepreneurial experience in internet marketing. He previously served as a senior product manager at iResearch Consulting, taking part in the development and improvement of multiple big-data products; he later became operations director at Blue Lion and chief strategy expert at Xinshu Network, advising several large internet companies on search marketing. He then made another entrepreneurial move in e-commerce, building a company from 0 to a team of one hundred and winning honors as one of Alibaba’s Top Ten Online-Goods Brands and Top Ten Global Online Merchants. He has also taught entrepreneurship at several universities and served as a guest lecturer at Taobao University—a quintessential hands-on practitioner.
In 2016, he founded Deci Mobile, focusing on mobile overseas search marketing (ASM), and was among the earliest Chinese companies to provide specialized services in this field, with clients spanning gaming, e-commerce, finance, and other internationally expanding industries. In 2024, he returned to the entrepreneurial starting line and founded DeerAPI, a service provider that covers 500+ large models worldwide and offers AI application developers an aggregated model-access platform.
He chose a role that can look like that of a mere “utility player”: building an aggregation platform for large-model APIs. There is no dazzling self-developed model and no slogan about a “cognitive revolution,” only one promise—“help developers call models faster and more reliably.” It may not sound glamorous, but it is exceptionally real. This is precisely the reality most AI entrepreneurs cannot avoid today: the market is already intensely competitive, models are becoming homogeneous, and the players that make the user experience solid have a chance to survive.
What DeerAPI does, in essence, is lower the barrier to “accessing AI.” It packages the APIs of almost every major model you can think of—GPT, Claude, Grok, Midjourney, Suno, and others—within a single platform. One registration, one top-up, and one set of calling standards turn a formerly cumbersome model-access process into a straightforward, plug-and-play experience. This “aggregation” is not a technological breakthrough but a judgment about value: saving developers time gives individuals and small teams more room to build.
Li Jinglin’s understanding of the “individual era” does not borrow from grand narratives, either. He talks about the independent webmasters of 2006 and the solo entrepreneurs who survived through independent cross-border e-commerce sites. Individuals were already breaking through boundaries then, he says; today’s AI simply fills their “capability gaps” more completely. Cannot write copy? Cannot edit video? Cannot code? AI can now fill all those gaps. The constraint on individuals is no longer skill, but willingness and execution.
He even says bluntly that today’s “individual + AI” combination is not a “lone hero” but a “one-person unit armed to the teeth.” This is not a romantic fantasy but a pragmatic judgment: AI is redrawing the boundaries of individual capability and genuinely changing how productivity is distributed. DeerAPI is like the logistics corps delivering ammunition to that unit—out of the spotlight, but indispensable.
Li Jinglin’s candor also extends to his view of the industry’s current state. He admits that he does not have the bandwidth to explore new paradigms such as Agentic AI, nor the capital to make a huge bet on an uncertain future. “We are still too small,” he says. “We are working hard to survive.” The phrase “stay alive” gives voice to the shared feeling of so many AI entrepreneurs. In this marathon-like technological transformation, survival already means winning half the battle.
His view of the core challenge in AI entrepreneurship is equally plainspoken: it is not whether you can “build a tool,” but whether you have “found the right people, solved the right pain, and can defend your ground.” In today’s environment, rushing an AI project into existence is all too easy. The hard part is whether you have truly solved a core problem for a specific group and whether you can prevent the solution from being copied. Without industry data or a defensible moat, a mere model wrapper will not go far.
When discussing “how individuals can capture the next wave of opportunity,” he speaks with an almost contrarian mindset. He does not urge you to improve your skills or preach advanced concepts. His advice is simply: do not overthink it—keep the project alive first. Profitability comes first; tools, skills, and systems come second. If you cannot protect today’s livelihood, tomorrow’s dream is empty talk. Instead of debating on paper, start doing what you want to do immediately. That capacity for execution is an individual’s real protection.
Li Jinglin and DeerAPI do not aspire to become the brightest star in the AI world, but they are doing something worthy of respect: amid an AI landscape dominated by giants, they are building a viable path for countless “non-giant” developers, individuals, and small teams to survive and get things done. It is not flashy, but it matters.
In this era, it is all too easy to be captivated by grand narratives such as “large models, AGI, and cognitive intelligence,” as though nothing has value unless it talks about the future. Li Jinglin reminds us that another narrative also deserves to be heard—one about the courage to survive, the resilience of individuals, and how tools can genuinely serve people.
In this era, every individual who manages to stand firm represents a victory. DeerAPI may be a sharp, practical blade in this battle for individual agency.
Selected Interview Q&A
Q1: Could you briefly introduce yourself and DeerAPI? What is the company’s core positioning?
Jingling: I am Jingling, the founder of DeerAPI. DeerAPI is a global API aggregation platform that brings together 500+ large models from around the world. Our core positioning is to help developers access all mainstream AI models in one place, including APIs for GPT, Claude, Grok, Midjourney, Suno, and many others. Our service offers exceptionally low prices, supports extremely high concurrency, guarantees 99% SLA stability, and enables rapid access.
Q2: What are DeerAPI’s principal products and solutions today, and which mainstream models do they cover?
Jingling: Our main offering is a one-stop large-model API access service designed to solve developers’ pain point of having to integrate with multiple model sources. We cover virtually all mainstream models on the market, which can broadly be grouped as follows:
Language models: These mainly include OpenAI (the GPT series), Claude, Gemini, Grok, Deepseek, and others.
Image models: These mainly include Midjourney, Stable Diffusion, Flux, Gemini-2.5-flash-image, GPT-4o-image, and others.
Video models: These mainly include Luma, Runway, Kling, Pika, and others.
Music models: These mainly include Suno, Uido, and others.
We have already become the preferred discounted API access platform for more than one hundred AI applications expanding overseas.
Q3: As a founder, what was your original motivation for creating DeerAPI, and what continues to drive you to work deeply in AI?
Jingling: My original motivation and vision were to help more AI applications and AI developers successfully “compete worldwide.” Several things drive me to keep working in this field: first, the passion of entrepreneurship itself; second, a firm belief that AI has vast prospects; and finally, the most practical factor—the pressure to survive. Combined with my past experience, these factors motivate me to work hard to stay alive.
Q4: What do you see as the most significant breakthrough in generative AI today, and how does DeerAPI apply these technologies internally?
Jingling: I believe the biggest breakthrough is that models are now evolving at a pace measured in “days.” You may think AI cannot replace people yet, but the speed of its evolution is astonishing.
For us, because DeerAPI itself is a platform for accessing large-model AI APIs, we naturally make extensive use of all kinds of large models in our daily work to improve the company’s own operating and R&D efficiency.
Q5: As AI technology iterates rapidly and competition at the model layer—for example, OpenAI and Claude—is intense, how does DeerAPI, as an intermediary layer, balance innovation with implementation?
Jingling: As an intermediary layer, when upstream model providers compete so intensely, we can only compete intensely alongside them. Homogeneity at the model layer is quite severe today. For us, the challenge is to find a balance among technological innovation, operating efficiency, and the team’s ability to execute under pressure.
Q6: Has DeerAPI begun preparing for emerging technological paradigms such as Agentic AI?
Jingling: We have not explored that area yet. Our priority right now is to do the work immediately in front of us well. The company’s business is still too small, so our first task is to find a way to survive.
Q7: The theme of this interview is “Creating Intelligence | The Individual Era.” How do you interpret it?
Jingling: The “individual era” is not actually a new term. As early as 2006 and 2009, many “individual pioneers” had already emerged, including independent webmasters and entrepreneurs running foreign-trade businesses on their own.
The AI era now greatly amplifies the power of individuals. In the past, people working alone were always held back by stages they were not good at—for example, not knowing how to write copy, shoot good video, or handle technical development. AI can now fill all of those “gaps.”
That is why I believe an individual in the AI era is not “fighting alone,” but is instead a super-combination of “one person + AI.” This combination delivers greater operating efficiency and productivity. What people call the “rise of the individual” is, in essence, AI giving individuals more powerful “weapons.”
Q8: How exactly does DeerAPI empower “individual creators” or “small teams,” and which concrete pain points does it solve?
Jingling: In the past, developers who wanted to access different AI models faced a great many hassles during development. They had to find each official source separately, spend large amounts of time working through the documentation for different platforms, top up accounts through each official channel, and coordinate the adaptation of different interfaces.
As a one-stop aggregation platform, DeerAPI exists to solve this pain point. Developers no longer need to visit multiple platforms. They only need to top up their account on DeerAPI, and within 1 minute they can access almost any model on the market that they want to use.
The result is an exponential increase in development efficiency—preparatory work that once took hours or even days can now be completed in 1 minute. People can devote their valuable energy to core feature development and innovation instead of wasting it on cumbersome “connection work.”
Q9: For “AI creators” (AI Creators) in your field, what are the greatest opportunity and the toughest challenges today?
Jingling: The AI industry is already intensely competitive. More specifically, there are two major challenges:
The first is finding a niche market and achieving PMF (product-market fit). You must be clear about the specific group you serve. Is their “pain point” something only you can solve precisely? If you fail to identify this “narrow entry point,” either nobody will use the product or it will quickly be displaced.
The second is whether your application-layer barriers and moat are strong enough. Can someone else copy what you have built easily? Is your advantage proprietary industry data, a self-developed algorithm, or the ability to integrate deeply with the workflows of a particular industry? If you simply call a general-purpose model and wrap a shell around it without core capabilities, you will be eliminated easily amid the intense competition.
Q10: Looking ahead over the next 1–3 years, what advice would you give individuals or small teams seeking to capture the benefits of AI?
Jingling: My advice is very practical: there is no need to make the path overly complicated. The central task is just one thing—solve the problem of “staying alive” first.
Do not begin by devising complex strategies or advanced modes of thought. For individuals and small teams, the most practical step is to find a product that can generate profit—even if it is only a small tool or service—and first establish stable operations.
Once you have stabilized, you can gradually use tools such as AI to improve your skills. In the early stage, however, you must focus on the core actions that “make money.”
I have also found that competition among individuals and small teams is not about “who has the better strategy,” but “who executes more effectively.” Their capabilities are broadly similar; those who ultimately break through are often the people who remain steady, dare to implement, and persist in doing small things well. In short: first secure “profitability and survival,” then add “skills and tools,” and rely on “execution” throughout.