
Minimax
MiniMax (Chinese: 上海稀宇科技有限公司, Shanghai Xiyu Technology Co., Ltd.) is a Shanghai-based AI company founded in December 2021 by a group of computer vision researchers who previously worked at SenseTime. The company's name is drawn from the minimax algorithm in game theory, which describes minimizing potential losses in a worst-case scenario while maximizing potential gains. Key founders include Yan Junjie (CEO), Yang Bin, and Zhou Yucong.
About
MiniMax (Chinese: 上海稀宇科技有限公司, Shanghai Xiyu Technology Co., Ltd.) is a Shanghai-based AI company founded in December 2021 by a group of computer vision researchers who previously worked at SenseTime. The company's name is drawn from the minimax algorithm in game theory, which describes minimizing potential losses in a worst-case scenario while maximizing potential gains. Key founders include Yan Junjie (CEO), Yang Bin, and Zhou Yucong.
Its first product, Glow, launched in October 2022, let users create virtual characters with backstories to chat with, and gained over 5 million users within four months. Early funding came from game developer miHoYo, and in March 2024 Alibaba Group led a $600 million financing round that valued the company at $2.5 billion.
#Products and Models
MiniMax builds multimodal foundation models spanning text, image, video, speech, and music, powering both consumer apps and an enterprise/developer platform:
- MiniMax-M2 / M3 series — its flagship large language models, used for text generation, reasoning, and coding
- Hailuo AI — a video-generation service that competes with tools like Runway and Sora
- MiniMax Audio / Speech and Music models — for voice and music synthesis
- Talkie / Xingye — AI companion/character chat apps (an evolution of the original Glow concept)
- MiniMax Agent — an agentic product line
Its M3 model reportedly scored 0.37 on one benchmark, slightly behind Anthropic's Opus 4.7 (0.42) and GPT-5.5 (0.39), but ahead of other competing models, and the company has also been iterating rapidly on its M2 series through mid-2026.
#Market Position and "AI Tigers"
MiniMax is considered one of China's so-called "AI tigers" — a cohort of startups building large language models to compete with U.S. AI giants like OpenAI — sometimes more specifically labeled one of the "Six Little Tigers" (六小虎) of Chinese AI.
As of its IPO prospectus, the company had served over 200 million cumulative users across more than 200 countries and regions. Its revenue reached $53.4 million for the nine months ending September 30, 2025, up about 174% year-over-year, though it posted a net loss of $512 million in the same period, driven mainly by revenue from subscriptions and in-app purchases.
#IPO and Financials
MiniMax went public in a landmark listing:
- It listed on the Hong Kong Stock Exchange on January 9, 2026, becoming the second major Chinese LLM developer to go public, one day after rival Zhipu AI
- The IPO raised HK$4.8 billion (about $620 million), and shares closed up 109% on the first day of trading at HK$345, versus an offer price of HK$165
- This debut valued the company at roughly $13 billion and set what MiniMax describes as a global record for the shortest time — about four years — from founding to public listing
- Its 2025 IFRS net loss was much larger, at $1.87 billion, driven mainly by non-cash fair-value losses on preferred shares as the company's valuation rose, while its adjusted non-IFRS net loss was $250.9 million
- By February 2026, annualized recurring revenue had surpassed $150 million, and Bloomberg later reported annualized revenue more than doubled to at least $300 million by late May 2026 ahead of a new flagship model rollout
The IPO also came against a broader backdrop: Chinese AI firms have been ramping up fundraising to compete with U.S. rivals while navigating Washington's export controls on advanced AI training chips.
Models

MiniMax: MiniMax M3
MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...

No reviews yet — be the first to share your experience.