A decade of elite training and one bold career gamble have produced one of the most striking company origin stories in China’s AI era. Yan Junjie, a Tsinghua and Chinese Academy of Sciences–trained researcher who once interned at Baidu’s deep learning institute, has built MiniMax into a publicly listed artificial‑intelligence firm whose market value briefly eclipsed that of Baidu in March 2026. The leap from intern to founder of a HK$3.5–3.8 trillion company took roughly four years from MiniMax’s founding in 2021.
Yan’s trajectory mixes orthodox credentials with atypical risk‑taking. Raised in Henan and steeped in mathematics and computer vision research, he held senior research and executive posts at SenseTime after finishing his doctoral work, enjoying equity incentives that might have produced early wealth if he had stayed for the company’s listing. Instead, a personal episode in 2021—helping an elderly relative record memories—prompted him to leave a cushioned path and found MiniMax, with a stated mission to make AI useful to ordinary people rather than just to specialist labs.
MiniMax positions itself as an AGI aspirant focused on multimodal models that handle text, speech, image and video. Since its incorporation it has raised about $1.5 billion across seven funding rounds from backers that include Alibaba, Hillhouse and prominent game and consumer‑tech investors. The firm lists a suite of consumer products such as a leading video‑generation tool and a conversational assistant with daily usage approaching social‑media norms, alongside an API platform and bespoke enterprise solutions for finance, law and education.
The company’s public market debut on Hong Kong’s Main Board on January 16th was emphatic: an IPO price of HK$165 per share, an oversubscribed retail tranche by more than 1,800 times and a first‑day jump of roughly 109 percent. The share price climbed rapidly through January and February and by early March valuations put MiniMax in the same league as long‑established internet giants—enough, for a time, to exceed Baidu’s market capitalisation on paper.
Three operational features stand out behind MiniMax’s momentum. First, an export‑oriented business mix: more than 70 percent of reported revenue comes from overseas consumer users, which reduces single‑market concentration. Second, aggressive cost and compute management: MiniMax advertises short chip rental cycles and a policy of training on the latest accelerators every two years to squeeze efficiency from cloud and hardware partners. Third, product stickiness—video and conversational offerings with long daily engagement metrics and a reported user base north of 212 million as of September 2025.
The MiniMax story is also a reminder that China’s AI ecosystem has become a global talent incubator. Baidu, which hired many leading researchers in the 2010s, is newly cast as a kind of academy; alumni include not just Yan but figures who later helped develop major Western models. Dario Amodei, for example, worked briefly in Baidu’s Silicon Valley lab before moving to Google and then OpenAI, later founding Anthropic. His early observations on model scaling would become central to the industry’s architecture choices.
The transnational paths of these founders highlight both opportunity and friction. Anthropic’s rise in the United States culminated in high‑value defence work and a public row with the U.S. Department of Defense over how its models are used—an episode that underlines the governance and geopolitical risks at the intersection of advanced AI, national security and commercial ambitions. For Chinese rivals like MiniMax, global traction brings revenue and influence but also scrutiny from customers and regulators abroad.
MiniMax’s ascent therefore matters beyond a headline valuation. It illustrates how startup dynamism, plentiful private capital and a pipeline of experienced researchers can challenge legacy technology firms in a compressed timeframe. The result is faster product iteration and intensified competition for talent and compute. The risks are clear as well: a heavy reliance on foreign consumer markets, exposure to international regulatory shifts, and an industry‑wide debate about safety, alignment and the exportability of powerful models. How MiniMax navigates those headwinds will be a bellwether for China’s broader AI ambitions.