Worth Buying: Developing the 'Worth Buying Consumption Content Large Model'


Chinese AI startup DeepSeek plans to conduct its first large-scale financing round, aiming to raise about 50 billion yuan (approximately 7.4 billion US dollars), with an estimated valuation of 350 to 400 billion yuan after the funding. The company has gained global attention with its V3 and R1 models, changing the perception of China's capabilities in large model development.
Zhipu AI has launched its IPO on the Science and Technology Innovation Board, planning to raise 15 billion yuan. It is less than five months since its listing on the Hong Kong stock market in January. The company is using the "A+H" dual-channel financing to broaden its funding sources and consolidate its leading position in the large model field.
Tencent Cloud's Intelligent Agent Development Platform announced a significant price reduction for the DeepSeek V4 series large models starting June 3, with the flagship DeepSeek-V4-Pro model seeing a 75% decrease in both inference input and output costs, aiming to lower innovation barriers for enterprises and developers while reducing daily invocation expenses.....
Moonshot (Kimi) is advancing its Hong Kong IPO, working on dismantling the VIE and red-chip structures, dissolving offshore companies under its Cayman parent company to clear regulatory obstacles. As one of the 'AI Six Dragons', this move marks a critical phase in its listing process.
Google launched its new large model Gemini 3.5 Flash at the I/O 2026 conference, breaking the AI speed record. The model features top-level agents and programming capabilities, with performance comparable to flagship models, while maintaining the high-speed response advantage of the Flash series. It is now freely available to users worldwide, further lowering the usage barriers and driving industry transformation.