$TENCENT (00700.HK)$ On February 2, the subsidiary TENCENT Cloud announced that the DeepSeek-R1 large model can be deployed to TENCENT Cloud HAI (TENCENT Cloud High-Performance Application Service) with one click, allowing developers to access it in just 3 minutes.

Tencent Cloud stated that, simply put, through HAI, developers can avoid tedious steps such as buying cards, installing drivers, configuring networks, configuring storage, setting up environments, installing frameworks, and downloading models, and can call the DeepSeek-R1 model in just two steps.

Previously, according to a news release from Huawei Cloud's official WeChat account on February 1st, the open-sourcing of DeepSeek-R1 has attracted global users and developers' attention. After continuous efforts from the Silicon-based Flow and Huawei Cloud teams, both parties have jointly launched and put online the DeepSeekR1/V3 inference service based on Huawei Cloud's Ascend Cloud Computing Service.

In recent days, DeepSeek has frequently surged to the trending searches. It is reported that AI giants like NVIDIA, Microsoft, and Amazon have announced their integration with DeepSeek. Additionally, perhaps pressured by DeepSeek, OpenAI has urgently launched the next-generation inference model o3-mini and opened the inference model to free users of ChatGPT for the first time.

According to reports from a US market observation website, NVIDIA's market cap has shrunk by 552 billion dollars over the past week.

DeepSeek, officially known as Hangzhou DeepSeek AI Basic Technology Research Co., Ltd., was established on July 17, 2023. It is an innovative technology company focused on developing advanced large language models (LLM) and related technologies.

On December 26, 2024, DeepSeek launched V3 (comparable to GPT4o). Due to its outstanding performance, the user base quickly began to grow. After the release of R1 (comparable to GPT-o1) on January 20, it further expanded its reach. According to data from Diandian, DeepSeek's download ranking in the USA rose rapidly from 201st on January 22 to the top rank on January 27; according to Qimai data, as of January 30, DeepSeek ranked first in download rankings in 168 countries.

On January 24, Marc Andreessen, founder of the prominent investment company A16z, mentioned in a post that Deepseek-R1 is one of the most astonishing and impressive breakthroughs he has ever seen, and it is also open-source, a gift to the world. The most provocative comment came from Alexander Wang, founder of ScaleAI, who said that while the USA might have been leading the AI race against China for the past decade, the release of DeepSeek's AI large model could 'change everything.'

DeepSeek recently released its latest open-source model DeepSeek-R1, achieving performance close to that of GPT-o1 developed by OpenAI at a lower cost. This advancement breaks the long-standing path dependence of the Global AI industry on merely stacking compute power, and its impact extends to the Capital Markets.

Editor/danial