摘要：Chinese President Xi Jinping attended a symposium on private enterprises in Beijing on February 17, with prominent attendees from

TMTPost -- Chinese artificial intelligence (AI) startup DeepSeek on Tuesday dismissed rumor about rollout of its next generation reasoning model R2 next week.

Credit：China Daily

Chinese news media outlet Sina Technology cited a source Tuesday that DeepSeek is going to release R2 on March 17. That is a “false news,” DeepSeek later that day responded.

Reuters almost two weeks ago reported DeepSeek is accelerating the launch of R2. It learned from people familiar with DeepSeek that the company previously planned to release the successor of R1 in early May, but now wants it out as early as possible. The Hangzhou-based company was reported to hope the new model will produce better coding and be able to reason in languages beyond English.

Chinese President Xi Jinping attended a symposium on private enterprises in Beijing on February 17, with prominent attendees from the business community including Liang and Alibaba Group founder Jack Ma. Xi delivered a speech at that symposium, urging healthy, high-quality development of private sector.

“On the new journey of the new era, the private economy has broad development prospects and immense potential, and it is the perfect time for private companies and entrepreneurs to fully demonstrate their prowess,” the state news agency Xinhua quoted Xi’s speech. “It is imperative to unify thinking, strengthen confidence, and promote the healthy and high-quality development of the private sector.”

DeepSeek has not raised any outside funding since its establishment in 2023. Part of the reason is that profits of its parent Chinese quant hedge fund High-Flyer has been able to meet the funding need. Liang said in 2023 that money had never been the problem for DeepSeek while bans on shipments of advanced chips are the problem. DeepSeek has experienced outages and malicious attacks amid overwhelming popularity since late January. It is also facing increasing scrutiny and restrictions from overseas authorities owing to concerns including data security. Despite these headwinds, Liang has not suggested the company needs outside funding yet.

Alibaba was said last month to plan $1 billion worth of investment in DeepSeek for 10% stake in the startup. But the Alibaba insider soon denied the report, calling it “fake news”. Some people in the industry believe that as computing power continues to be in short supply, it is time for DeepSeek to seek the outside funding since the firm is one of a few large model developers that has not yet initiated its first funding round and is delivering leading models comparable to OpenAI’s at a lower cost.

The Information reported on February 19 that DeepSeek is weighing outside fundraising due to the need for more AI chips and servers to deal with fast-growing usage and support model development. Many investors including Alibaba Group and Chinese state funds like China Investment Corp., the country’s sovereign wealth fund, and the National Social Security Fund, have expressed their interests in financing DeepSeek, according to the report.

The report said executives at DeepSeek and High-Flyer are talking about whether focus of the Chinese AI firm should pivot from research to building a business generates revenue and eventually profit. The report suggested part of the reason that DeepSeek is reluctant to raise money from outsiders is it wants to avoid pressure from investors to commercialize its products too quickly. It noted DeepSeek founder and CEO Liang Wenfeng is still cautious about outside fundraising.

DeepSeek insiders a day after described the aforementioned report as false.

来源：钛媒体