I am deeply fascinated by the deepseek free app, a Chinese reasoning chatbot, DeepSeek, which has become the number one free app downloaded by Apple and Android Play Stores. Deepseek recorded downloads on Apple and Android Play stores have hit 16 million and 10 million as of January 23rd and January 26th, respectively. A combination of factors has made the deepseek app most appealing as a digital search tool, namely, a competitive search engine compared to its rival apps such as ChatGpt, Google Gemini, LLama, etc., and charge nothing! Other apps of similar fraternities charge user fees, but DeepSeek raises the question of how DeepSeek clears her bills? I asked DeepSeek herself by turning myself into a shamus to get to the bottom of who pays Deepseek. This question is paramount to understand if whoever pays the piper calls the tune, implying that security concerns may be grave if the Chinese government offsets its bills. This discussion peeks at these complex financing issues and offers some tentative observations.
DeepSeek, like many other companies offering free services, likely employs a combination of strategies to finance its operations without directly charging users. Here are some nine common approaches:
- Advertising: One of the most common ways to generate revenue is through advertising. DeepSeek could display ads to users in future and charge advertisers based on impressions (views) or clicks. So far, this model has not been invoked, so no money has trickled down from this frontier. It could be an option once the number of users grows to half a billion. DeepSeek may decide to follow Tiktok’s revenue-sharing model, which will definitely encourage users to contribute to her development. How that will work is anybody’s guess, but making queries available for access to others and generating views and likes could be one way of determining how much users can be compensated for their efforts. This will be a better improvement over other digital libraries like Wikipedia, which does not monetize and compensate its contributors. Wikipedia’s chief source of income is to urge users to voluntarily contribute money monthly, quarterly or annual plans based on philanthropic gestures to keep it afloat.
- Freemium Model: While the core services are free, DeepSeek might offer premium features or additional services in the future that users can pay for. This model allows users to access basic functionalities for free while generating revenue from those who need more advanced features. However, the problem with this approach is the alienation of the majority of users who are vital to building a digital data global store.
Deepseek’s latest R1 model beat OpenAI’s O1 on multiple reasoning benchmarks. High-Flyer fully funds Deepseek, which has no plans to raise funds. The company focuses on building foundational technology rather than commercial applications and has committed to open-sourcing all of its models. It has also single-handedly kicked off price wars in China by charging very affordable API rates.
Before the Deepseek era, CEO Liang Wenfeng’s leading venture was High-Flyer (幻方), a top-four Chinese quantitative hedge fund last valued at $8 billion. Deepseek’s strategy is grounded in its ambition to build AGI.
Unlike previous spins on the theme, Deepseek’s mission statement does not mention safety, competition, or stakes for humanity but only “unravelling the mystery of AGI with curiosity”. Accordingly, the lab has been laser-focused on research into potentially game-changing architectural and algorithmic innovations.
My conclusion: since the owners of Deepseek are loaded and their business model is cost-effective, then short-term funding is not an issue since it is a digital lab that is collecting and analysing digitally available data that it harvests and may sell or turn it into a profitable venture through other generative AI ventures. Hence, Deepseek is investing in not charging user fees in order to widen her appeal and, in the process, have a larger audience to assist her in developing and managing possibly the largest workshop in the world that would outpace those of the same fraternity hamstrung by subscription fees.
- Data Monetization: With user consent and in compliance with privacy laws, DeepSeek could analyze user data to gain insights and sell these insights to third parties, such as market researchers or advertisers. This is often done anonymously and aggregated to protect user privacy.
DeepSeek’s primary data monetization model appears to be based on providing free access to its advanced AI models, like DeepSeek R1, while potentially generating revenue through partnerships, research collaborations, and future development of specialized, potentially paid features; this strategy contrasts with other AI companies like OpenAI, which primarily monetize their models through subscription fees, as DeepSeek aims to gain market share by making its technology widely accessible and open-source.
- Partnerships and Sponsorships: DeepSeek might form partnerships with other companies or secure sponsorships. These partners or sponsors could pay DeepSeek to integrate their services or products within the platform. By making their technology widely available, DeepSeek aims to build a large community of users and developers who can contribute to further advancements and potentially generate new applications.
- Affiliate Marketing: DeepSeek could earn commissions by referring users to other services or products. When users click on affiliate links and make purchases, DeepSeek receives a percentage of the sale. This model is still in its infancy and had yet to be deployed when I prepared this article.
- Enterprise Solutions: DeepSeek might offer specialized services or licenses to businesses and enterprises. These solutions could include advanced analytics, custom integrations, or dedicated support, which are charged at a premium. Deepseek is also offering customized versions of its AI models with additional features or support for specific industry needs. Regarding data licensing, DeepSeek is potentially selling access to specific datasets used to train its models, depending on data privacy regulations.
- Crowdfunding and Donations: Some platforms rely on user donations or crowdfunding to support their operations. DeepSeek could encourage users to contribute voluntarily to keep the service free for everyone. Wikipedia currently uses this model.
- Investment and Venture Capital: In the early stages, DeepSeek might rely on investments from venture capitalists or angel investors. These investors provide funding in exchange for equity, betting on the company’s future profitability. So far, I have seen no evidence that DeepSeek was soliciting this kind of financing. Perhaps DeepSeek is unwilling to give up her say over how the company will be run in such tender stages of development. Investors love to have a say over major policy decisions, which may conflict with pioneers’ originality and creativity.
- Government Grants and Subsidies: Depending on the nature of the service, DeepSeek might qualify for government grants or subsidies aimed at supporting innovation, education, or other public benefits. The downside of depending on governments’ subsidies is that strategic interests may collide, compromising privacy and data security or constraining the app from achieving global acceptability, a key ingredient to fostering digital data generation with a global outlook. Once other governments see an unhealthy relationship exists between DeepSeek and the Chinese government, they may feel threatened, leading to regulatory bans imposed on the app under the motif of safeguarding national security.
By leveraging one or more of these strategies, DeepSeek can sustain its operations and continue to offer free services to its users.
My closing remarks.
Deepseek is a digital lab that affords her majority of clients zero user subscriptions to attract them to contribute to building the most in-depth, powerful, intelligent AI and accessible social reasoning platform. While its owners are multi-billionaires and may not have to worry very much about the running costs of this enterprise in the long run, Deepseek will generate profits by powering other generative AI with marketability value from the digital data harvested in this app. Deepseek is analogous to building a water dam to collect water that will eventually be used to irrigate farms and provide drinking, cooking and cleaning water for domestic and industrial use. Rather than seeking to collect subscription fees upfront, the Chinese entrepreneur Liang Wenfeng opted to absorb user costs in order to amass user data to train other generative AI for commercial purposes. In this way, whatever losses incurred from foregone subscription fees are quickly recovered elsewhere, as demonstrated earlier.
Whoever uses Deepseek in search of answers automatically becomes a small investor in future developments and is paid upfront by not clearing subscription fees. That way many willing small investors join the enterprise for a better cause. Whatever queries that are not satisfactorily answered mean future developments will accommodate those shortfalls, making the app even more powerful and precise. While the owners are loaded, they must cooperate with the Chinese authorities to ensure China’s national security is not compromised. By doing so, other countries feel that Deepseek may be spying on them, and they logically suspect that Deepseek is the Chinese spying machine.
It may dilate why some of China’s unfriendly countries have begun imposing restrictions on the app out of the trepidation of capitulating into the Chinese microscopic lenses that are carefully predicated for military unfair advantages or powering tools of propaganda and misinformation. This expounds why some countries have banned the app in government phones and other electronic devices out of fear of being hacked and spied upon through this app. So far, there is no evidence to support these perceptions. Foreign regulatory risks may constrain the development and accessibility of this remarkable digital venture. Those regulatory risks are formidable, as Apple is now discovering.
Digital platforms are now braving a new form of censorship, as witnessed in the UK, where His Majesty the King’s government has ordered Apple to furnish her with unreasonable access to its iCloud services in order to spy on her perceived adversaries under the guises of protecting children against paedophiles, trafficking gangs and similar societal vices. How the saga unfolds and concludes will have a far-reaching impact on how digital platforms perform and conduct themselves in the long run.
Read more about Chinese startup deep seek unravels tech industry
