With just a $2.19 payment to use what seems like its most advanced model, the Chinese AI-chatbot app DeepSeek is delivering an upset for Artificial Intelligence companies in the United States whose process of enabling machines to assist humans in their lives is obviously way too costly and that has meant at the moment the shifting attention away from them, but is the new alternative from East Asia to be trusted wholeheartedly?

First is to understand what the Hangzhou-based startup in China DeepSeek can offer compared to other competitors in the AI space like ChatGPT founded by OpenAI in December 2015.

So far, the former has received exceptionally high interest, leading to the platform experiencing outages since launching the DeepSeek-R1 model eight days ago on 20 January 2025. It is the company’s second approach to computers finding better ways to help humans work better.

Even though the Chinese version obviously has limitations when outsiders from the Western world look at it, what it offers is much cheaper. For example, DeepSeek-V3 required less than $6 million worth of computing power from Nvidia H800 chips, although the AI platform can also be said to be pro-China in its views.

Does that make this particular AI platform what should be considered honest or not? What is something remarkable or obscure that techies looking for cheaper assistance in Nigeria to look out for?

ALSO READ: Megacity of Lagos closer to realising its own AI destiny where etiquette can’t be stressed enough

What DeepSeek does is that it presents itself as a promising yet controversial player in the competitive AI chatbot market. It recently created a stir with its AI-chatbot app in the Apple store for iPhone users by offering advanced models at a fraction of the cost compared to U.S-based AI companies.

The latest model, DeepSeek-R1, significantly contributed to its popularity. So what sets this new hack apart from the others that were supposed to be way advanced because of the billions of dollars they were going to get from Oracle, Nvidia, and Microsoft?

Who is the best?

One of DeepSeek’s most striking features is its cost efficiency. As mentioned earlier, the development of DeepSeek-V3 required less than $6 million worth of computing power from relatively outdated Nvidia H800 chips5. Compared to giants like OpenAI, which spend billions on AI model development, DeepSeek’s approach is significantly cheaper.

Also, despite using less advanced hardware, DeepSeek’s performance is competitive. Benchmark tests indicate that DeepSeek-V3 outperforms well-known models like Llama 3.1 and Qwen 2.5 while matching the capabilities of GPT-4o and Claude 3.5 Sonnet

. This impressive performance at a lower cost has unsettled some U.S. AI companies.

It has to be said that DeepSeek’s commitment to open-source models has democratised AI technology. Unlike several other AI firms that operate on a subscription model, DeepSeek has fully open-sourced its models under an MIT license, making them accessible for both commercial and academic applications. This move has propelled its popularity, especially among developers and researchers.

Based on data measuring its success, DeepSeek’s new R1 model shows impressive results, scoring 79.8% Pass@1 on AIME 2024—and it is performing better than OpenAI’s o1-1217.

In addition, the Chinese platform offers grooming to those still finding calculations difficult due to excelling greatly at mathematics and coding tasks. It is reported to have achieved a 97.3% score on the MATH-500 standard and performed much better than 96.3% of human participants in Codeforces coding competitions.

But for how long will trustworthiness and strict compliance with the People’s Republic of China’s views of itself and Taiwan, the island nation the Chinese leadership is ready to take by force be an issue?

DeepSeek’s noteworthy achievements and low-cost development model have not come without concerns. The app has already faced scrutiny regarding its data privacy policies.

Being a Chinese company, there are rising concerns that data collected could potentially fall into the hands of the Chinese state. Such apprehensions have precedent, as seen with apps like TikTok and companies like Huawei.

The platform’s transparency and openness can be a double-edged sword. While its open-source nature provides an opportunity for independent audits and checks, its pro-China rhetoric remains a point of contention. Users are urged to exercise caution and ensure personal data isn’t compromised while using the chatbot.

What users in Nigeria probably should consider

Tech enthusiasts and professionals in Nigeria should not overlook some of DeepSeek’s remarkable features and potential obscure pitfalls.

As one of its advantages, the reduced costs make advanced AI technologies more accessible to startups and individual developers in emerging markets like Nigeria. The platform’s impressive performance and reasonable API pricing could be incredibly beneficial for those seeking cost-effective AI solutions.

Potential cautions include the need to approach the app carefully regarding data privacy. Users should be well-informed about DeepSeek’s data collection, storage, and usage policies before heavily relying on it.

Ensuring that sensitive or personal data isn’t fed into the chatbot and watching for biased responses can safeguard user interests. Users in Nigeria, therefore, ought to be conscious of potential biases. Only if they carefully navigate these concerns will enthusiasts enjoy the features that have been highlighted.

ALSO READ: Digital economy minister Bosun Tijani on live TV defends his AI fixation as necessary

Ayodele Johnson

Ayodelé is a Lagos-based journalist and the Content and Editorial Coordinator at Meiza. All around the megacity, I am steering diverse lifestyle magazine audiences with ingenious hacks and insights that spur fast, informed decisions in their busy lives.